Emotion analysis-based sentiment intervention system for elderly online learning

Abstract

The invention discloses an elderly online learning emotion intervention system based on emotion analysis. The elderly online learning emotion intervention system comprises a recording module, a face recognition module, an emotion feature analysis module, a time emotion map drawing module, a behavior dynamic transfer module and an intervention strategy module. The recording module is used for automatically recording face videos in the online learning process of old people and acquiring face emotion data; and the face recognition module is used for carrying out statistical analysis on the facial emotion data of all old people. The invention belongs to the technical field of emotion intervention, and aims to solve the problems that in the prior art, facial expression data of old people in an online learning scene cannot be accurately captured, so that the accuracy of subsequent emotion recognition is not high, and effective intervention measures cannot be provided. The technical effects are that facial expression data of the old people in an online learning scene can be accurately captured, the accuracy of subsequent emotion recognition can be improved, and effective intervention measures can be provided.

Description

Technical Field

[0001] The present application belongs to the field of emotion intervention technology, and specifically relates to an emotion intervention system for elderly online learning based on emotion analysis. Background Art

[0002] Elderly age is a period of heightened emotional needs. Senior online learning aims to enhance their emotional experience and social participation. Online learning provides a natural setting for large-scale analysis of seniors' emotions. Emotions are a key focus in gerontology and senior online learning. Research has shown that online learning can significantly improve seniors' emotional well-being, enhance their subjective well-being, positively impact their mental health, and alleviate depressive symptoms.

[0003] The reason online learning for seniors can achieve such success is that, on the one hand, the acquisition of knowledge and skills increases opportunities for self-realization, and on the other hand, it provides avenues for social participation and serves as an important channel for emotional support. For seniors experiencing emotional deficits or exhibiting depressive tendencies, designing and implementing targeted educational interventions can help improve their mental health. Therefore, sentiment analysis plays a crucial role in gerontology and online learning for seniors, contributing to the precision of course design and providing a basis for interventions addressing negative emotions in the elderly. Analyzing seniors' emotions and implementing educational interventions based on affective computing technology are of contemporary significance for improving negative emotions in the elderly.

[0004] Facial expressions, as a natural channel for human emotional expression, are relatively authentic and easily accessible, making them a commonly used data modality in affective computing. Natural scenes learned online can objectively reflect the true emotional states of older adults, providing a convenient channel for capturing and analyzing their facial expressions on a large scale.

[0005] However, most of the current emotional intervention systems for elderly online learning are unable to accurately capture the facial expression data of the elderly in online learning scenarios, resulting in low accuracy in subsequent emotion recognition and an inability to accurately judge the emotional state of the elderly, thus failing to provide effective intervention measures. Summary of the Invention

[0006] This application provides an emotional intervention system for elderly online learning based on sentiment analysis, which aims to solve the problem that existing technologies are unable to accurately capture the facial expression data of the elderly in online learning scenarios, resulting in low accuracy of subsequent emotion recognition and thus inability to provide effective intervention measures.

[0007] An emotion intervention system for elderly online learning based on emotion analysis, including a recording module, a face recognition module, an emotion feature analysis module, a time emotion map drawing module, a behavior dynamic transfer module, and an intervention strategy module;

[0008] The recording module is used to automatically record facial videos of elderly people during online learning and obtain facial emotion data;

[0009] The face recognition module is used to perform statistical analysis on the facial emotion data of all elderly people and analyze the average frequency of facial emotions that appear in the elderly in the online learning scenario;

[0010] The emotional feature analysis module is used to analyze the structural features of the facial expressions of the elderly using a machine learning data mining method;

[0011] The time emotion map drawing module is used to draw a smooth broken line graph reflecting the emotional changes of the elderly over time, and the time emotion map drawing module includes a data preprocessing unit;

[0012] The behavior dynamic transfer module is used to perform a lag sequence analysis on the facial expression behavior of the elderly, obtain a behavior transfer frequency table and a Z-score score table, and derive the negative emotions of the elderly and their transfer rules;

[0013] The intervention strategy module is used to judge the richness of emotional behavior transfer of different types of elderly people according to the Z-score score table, and automatically match and push differentiated intervention content according to the richness of transfer.

[0014] Furthermore, the recording module has a recording duration of 1 hour during the online learning process, and allows interactive behaviors such as pausing and jumping back and forth during the online learning process.

[0015] Furthermore, the face recognition module is constructed based on FaceReader software, and the training database of FaceReader software contains facial expression data of the elderly and East Asians.

[0016] Furthermore, the machine learning data mining method is cluster analysis, which is used to calculate the sum of clustering errors of all samples of facial emotion data of the elderly.

[0017] Furthermore, the smooth line graph is drawn using the Python language and Matplotlib library using the facial expression recognition detailed data exported from the FaceReader software.

[0018] Furthermore, the data preprocessing unit is used to perform normalization processing and noise elimination processing on the facial emotion data.

[0019] Furthermore, the intervention strategy module includes a preset unit and a matching unit, and the preset unit is used to define the richness classification of emotional behavior transfer based on the Z-score score table.

[0020] Furthermore, the matching unit automatically identifies the emotion type of the elderly based on the preset emotion behavior transfer richness classification and accurately matches the corresponding intervention content.

[0021] Compared with the prior art, this application has at least the following beneficial effects:

[0022] This application, based on further analysis and research of existing technical issues, uses a recording module and a facial recognition module to accurately capture the facial expression data of seniors in online learning scenarios. Using facial data as a key information source, in-depth analysis reveals patterns in seniors' emotions. This pattern not only provides a new perspective for understanding the emotional state of seniors during learning but also helps inform appropriate emotional intervention strategies, thereby more effectively enhancing the effectiveness of emotional support. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 A schematic diagram of a module of an emotion intervention system for elderly online learning based on emotion analysis provided in one embodiment of the present application;

[0024] Figure 2 A schematic diagram of SSE value changes in an online learning emotion intervention system for the elderly based on emotion analysis provided in one embodiment of the present application; DETAILED DESCRIPTION

[0025] In order to make the purpose, technical solutions and advantages of this application more clear, this application is further described in detail below with reference to the accompanying drawings and embodiments.

[0026] Example 1

[0027] like Figure 1 and Figure 2 As shown, the present application provides an emotional intervention system for elderly online learning based on emotional analysis, including a recording module, a face recognition module, an emotional feature analysis module, a time emotion map drawing module, a behavior dynamic transfer module and an intervention strategy module.

[0028] The recording module automatically records facial videos of seniors during online learning and captures facial emotion data. Recording time is limited to one hour to minimize fatigue. Interactive behaviors such as pausing and skipping are allowed during the learning process, and facial expressions are recorded using a computer camera.

[0029] Online learning for the elderly was chosen as the research scenario because, on the one hand, online learning is an important and normal form of elderly education; on the other hand, online learning is less affected by human factors and can obtain more natural emotional expressions from the elderly; finally, the online method can realize the large-scale and automated collection of data, facilitating intelligent analysis.

[0030] The facial recognition module performs statistical analysis on facial emotion data from all elderly individuals, analyzing the average frequency of facial emotions expressed by them in online learning scenarios. This module is built on FaceReader software, which is based on Ekman's basic emotion theory and can analyze seven emotions with high accuracy. Chinese researchers have validated this system's ability to recognize facial expressions in Chinese people.

[0031] At the same time, the training database of FaceReader software contains facial expression data of the elderly and East Asians, which can effectively complete the emotion recognition task required by this study.

[0032] The emotion feature analysis module is used to analyze the structural features of facial expressions of the elderly using machine learning data mining methods. Among them, the machine learning data mining method is cluster analysis. Cluster analysis divides the research samples into different categories through mathematical methods, so that the similarity of samples within each category is as high as possible, while the similarity of samples between different categories is as low as possible. It does not pre-set the number and structure of clusters, but objectively classifies according to data features, which can effectively avoid human interference and make the classification results more consistent with the actual distribution of the data. K-means clustering is a typical unsupervised clustering algorithm. Using relevant libraries in Python, cluster analysis can be performed on facial emotion data of the elderly. The calculation formula is as follows:

[0033]

[0034] Among them, SSE (Sum of the Squared Error) is the sum of the clustering errors of all samples, i represents the i-th cluster, p is the sample point, m i is the mean of the samples in cluster i. Lower SSE values indicate better clustering results. As the number of clusters K increases, the sample divisions become finer, the clusters become denser, and the SSE decreases accordingly. However, when K is smaller than the actual number of clusters, increasing K causes the SSE to decrease rapidly. When K is larger than the actual number of clusters, the rate of decrease in SSE due to increasing K slows. Therefore, the turning point in the slope of the SSE curve represents the optimal K value.

[0035] Under the K-means clustering method, the corresponding SSE value of expression data changes with different K values.

[0036] like Figure 2 As shown in the figure, when K is greater than 5, the SSE value decreases slowly as K increases, so 5 is the optimal number of clusters.

[0037] The temporal emotion map drawing module is used to draw a smooth line graph reflecting the emotional changes of the elderly over time. By utilizing the detailed expression recognition data exported from the FaceReader software, the Python language and Matplotlib library are used to draw a smooth line graph reflecting the emotional changes of the elderly over time.

[0038] At the same time, the temporal emotion map drawing module also includes a data preprocessing unit, which is used to normalize and eliminate noise on the facial emotion data. The specific contents are as follows:

[0039] Because the total length of each elderly person's emotional recordings varied slightly, the facial emotion data was normalized to facilitate comparison of emotional changes across learners at different timescales. To eliminate data noise, the data was smoothed using an exponentially weighted moving average (EMA), giving higher weight to recent data to better capture data trends and make the data more interpretable.

[0040] The behavioral dynamic transfer module is constructed based on the interactive behavior analysis software GSEQ (General Sequential Querier). It is used to perform lagged sequence analysis on the facial expression behaviors of the elderly, obtain the behavioral transfer frequency table and Z-score score table, and deduce the negative emotions of the elderly and their transfer patterns, providing guidance for the subsequent design of effective educational intervention measures, thereby improving the emotional state of the elderly.

[0041] The intervention strategy module is used to judge the richness of emotional behavior transfer of different types of elderly people based on the Z-score score table, and automatically match and push differentiated intervention content based on the richness of transfer. The intervention strategy module includes a preset unit and a matching unit;

[0042] The preset unit is used to define the richness classification of emotional behavior transfer based on the Z-score score table.

[0043] For example, the emotional types of the elderly are classified into three categories according to the Z-score table, including anxious type elderly, happy type elderly and serious type elderly, and the corresponding intervention content is input.

[0044] The matching unit automatically identifies the emotional types of the elderly based on the preset emotional behavior transfer richness classification, accurately matches the corresponding intervention content, and pushes differentiated intervention strategies through the system.

[0045] In this emotional intervention system for seniors learning online, based on sentiment analysis, a recording module and a facial recognition module work together to accurately capture the facial expressions of seniors during online learning. Using facial data as a key information source, in-depth analysis reveals patterns in seniors' emotions. This pattern not only provides a new perspective for understanding the emotional state of seniors during learning but also helps inform appropriate emotional intervention strategies, thereby more effectively enhancing the effectiveness of emotional support.

[0046] Example 2

[0047] Thirty-two elderly individuals (8 men and 24 women) were randomly recruited from a senior citizen college in Beijing. The online learning platform used was the National Public Service Platform for Senior Education, developed by the National University for the Aged. The participants were selected to watch the 10-session video course "How Much Do You Know About Health," which has a different theme.

[0048] For example, topics like "Do I have any illness?" and "Who can die from a cold" were tailored to the learning preferences of seniors. Recordings were limited to one hour to minimize fatigue. Seniors were allowed to interact with the learning process, such as pausing and skipping forward and backward. Their facial expressions were recorded via a computer camera during the online learning process. Researchers provided technical support throughout the entire process, ensuring a smooth experiment and preventing any technical barriers to online learning from impacting the seniors' emotions.

[0049] The facial videos of the elderly were analyzed using the face recognition module, and the facial emotion summary data of some of the elderly are shown in Table 1.

[0050] Table 1 The duration of facial emotions of some subjects

[0051]

[0052] The facial emotion data of all elderly people were statistically analyzed, and the results are shown in Table 2.

[0053] Table 2 Facial emotion statistics of all subjects

[0054]

[0055] The average frequency of facial emotions expressed by older adults in online learning scenarios, from highest to lowest, was neutral (61.6%), sadness (15.2%), anger (7.5%), happiness (4.3%), disgust (3.0%), surprise (1.5%), and fear (0.2%). This indicates that neutral emotions are predominant among older adults. Furthermore, sadness was significantly more frequent than the other five emotions.

[0056] Clinical research shows that sad expressions are associated with depression severity. Depressed patients are significantly more likely to display sad expressions than the general population, and this rate increases with increasing depression severity. Therefore, the risk of depression in the elderly is higher than in the general population.

[0057] We further investigated the gender differences in the frequency of facial emotions among the elderly. The results of the independent sample T-test are shown in Table 3.

[0058] Table 3 Mean frequency of facial expressions in different gender groups

[0059]

[0060] Females showed more neutral expressions (65.4% > 50.2%) and disgust expressions (3.9% > 0.4%), while males showed more angry expressions (16.50% > 4.56%), reaching a statistically significant level.

[0061] Based on the facial expression data of 32 elderly people, cluster analysis method is used to analyze the structural characteristics of facial expressions of the elderly.

[0062] Based on the above sentiment computing technology and cluster analysis algorithm, the emotional characteristics of the elderly in the online learning scenario can be clustered into five types. Based on the average frequency of the six expressions in each group, excluding neutral, a radar chart was drawn, and the following results were obtained:

[0063] The first group of elderly people mainly expressed neutral emotions, and the average frequency of the other six types of expressions was less than 2%.

[0064] Especially in the other four groups;

[0065] While sadness levels were generally high among older adults, only this group showed a low frequency of sad expressions (1.39%), suggesting that this group exhibited less emotional fluctuations and maintained a relatively calm state of mind. This type, known as the calm type, accounted for 21.88% of the elderly participants.

[0066] The second group of elderly people showed a higher incidence of angry expressions (22.53%), far exceeding the frequency of angry expressions in the other four groups. Sad expressions were also more common (21.37%). This type of expression can be called the anxious type, accounting for 25% of the elderly subjects.

[0067] The average frequency of disgust expressions in the third group (14.3%) was significantly higher than in the other four groups, accompanied by a higher frequency of sadness (20.88%). Four other expressions, happiness (2.2%), surprise (0%), fear (0%), and anger (2.74%), were less common. This type of expression, which can be termed the boredom type, accounted for 15.63% of the elderly subjects.

[0068] The fourth group of elderly people displayed a higher frequency of sad expressions (17.04%), while the average frequency of the other five expression types was less than 5%. The scatter plot of the cluster results shows that this group differs from the boredom type primarily in that they display a lower frequency of disgust expressions. The serious type, which can be designated as the serious type, accounted for the highest proportion of the elderly participants, at 34.38%.

[0069] The fifth group showed a higher frequency of happy expressions (21%), while the average frequency of surprised expressions (7.2%) was significantly higher than in the other four groups. While some sadness expressions (14.9%) appeared, their frequency was lower compared to the other groups. Negative expressions such as fear (5.4%), disgust (0%), and anger (0%) were less common. This group showed a relatively high number of positive expressions, designated the happy type, which accounted for 3.13% of the elderly subjects, the lowest proportion.

[0070] The temporal emotion map drawing module is used to draw a smooth line graph reflecting the emotional changes of the elderly over time, and is used to display the emotional change characteristics of different types of elderly people over time.

[0071] In the smooth line graph of the emotional changes of calm elderly people, except for neutral emotions, the frequencies of other emotions are low and basically unchanged; the curve is flat and the fluctuation is small.

[0072] In the smooth line graph of the emotional curve of anxious elderly people over time, sadness and anger are two types of emotions that appear significantly more frequently, and the frequency of these two types of emotions has a slight upward trend over time; the curve is flat and has little fluctuation.

[0073] In the smooth line graph of the emotional changes over time of the boredom type of elderly people, sadness and disgust are two types of emotions that appear more frequently; the curve is not smooth and fluctuates.

[0074] In the smooth line graph of the emotional changes over time of serious elderly people, sad emotions obviously appear more frequently, while other types of emotions appear less frequently; the curve is flat and the fluctuation is small.

[0075] In the smooth line graph of the emotional changes over time of happy elderly people, multiple emotions coexist, but happy emotions are slightly higher than other emotions; emotions change in a wave-like manner with large fluctuations.

[0076] As shown in Table 4, a comparison table of coding letters and facial expressions was set up to facilitate subsequent research on the negative emotions of the elderly and their transfer patterns, and to provide guidance for the subsequent design of effective educational intervention measures.

[0077] Table 4 Comparison table of coded letters and facial expressions

[0078]

[0079] As shown in Tables 5 and 6, GSEQ analyzed the frequency and Z-score of emotional transfers among all the elderly participants in the study. "Initial behavior" refers to the first action, and "latent behavior" refers to the subsequent action. For example, the intersection of "Initial behavior" and "Lagged behavior" indicates that expression N was followed by expression SU.

[0080] Table 5 Frequency of facial emotion transfer among the elderly

[0081]

[0082] As shown in Table 6, statistically significant behavioral sequences were screened. If the adjusted residuals showed a Z-score > 1.96, the behavioral transition was significant. Overall, the Z-scores for H→SU, SU→H, SA→A, and A→SA were all greater than 1.96. The pattern of emotional transitions among older adults was found to be statistically significant: the four types of sequential transitions from happiness to surprise, surprise to happiness, sadness to anger, and anger to sadness, while the other types of sequential transitions were not significant. Furthermore, although the transition frequency between neutral expressions and other types of expressions was extremely high, the Z-scores were all less than 1.96, indicating that these behavioral transitions were not significant. This suggests that neutral expressions are a normal expression.

[0083] Table 6 Z-score table of facial emotion transfer sequence of elderly people

[0084]

[0085] We further explored the emotional transfer patterns of different types of elderly people and obtained significant transfer behaviors as shown in Table 7.

[0086] Table 7 Emotional behavior transfer and Z-score of different categories of elderly people

[0087]

[0088] The following can be concluded from Table 7:

[0089] (1) Anxious elderly people have rich emotional and behavioral transfers. There are significant bidirectional behavioral transfers between happiness and surprise, between sadness and anger, and between sadness and fear. In addition, there are significant behavioral transfers from happiness to disgust and disgust to surprise.

[0090] (2) Happy elderly people only have a significant behavioral shift from sadness to happiness.

[0091] (3) In addition to the significant bidirectional behavioral shift between happiness and surprise, serious elderly people also have a significant behavioral shift from sadness to fear.

[0092] The findings reveal that elderly individuals with anxious personalities exhibit behavioral transfers between negative emotions, with a bidirectional cycle between sadness and anger, and between sadness and fear, being absent in elderly individuals with other personality types. This discovery of behavioral transfer patterns between negative emotions in elderly individuals with anxious personalities is significant, demonstrating that negative emotions can be continuously transferred between them, leading to a repetitive, uninterrupted cycle of negative emotions and a cumulative state of negative emotion. Elderly individuals with anxious personalities are more susceptible to depression, and therefore require special attention and support.

[0093] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. An emotional intervention system for elderly online learning based on sentiment analysis, characterized by: It includes recording module, face recognition module, emotion feature analysis module, time emotion map drawing module, behavior dynamic transfer module and intervention strategy module; The recording module is used to automatically record facial videos of elderly people during online learning and obtain facial emotion data; The face recognition module is used to perform statistical analysis on the facial emotion data of all elderly people and analyze the average frequency of facial emotions that appear in the elderly in the online learning scenario; The emotional feature analysis module is used to analyze the structural features of the facial expressions of the elderly using a machine learning data mining method; The time emotion map drawing module is used to draw a smooth broken line graph reflecting the emotional changes of the elderly over time, and the time emotion map drawing module includes a data preprocessing unit; The behavior dynamic transfer module is used to perform a lag sequence analysis on the facial expression behavior of the elderly, obtain a behavior transfer frequency table and a Z-score score table, and derive the negative emotions of the elderly and their transfer rules; The intervention strategy module is used to judge the richness of emotional behavior transfer of different types of elderly people according to the Z-score score table, and automatically match and push differentiated intervention content according to the richness of transfer.

2. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The recording module records for 1 hour during the online learning process, and allows interactive behaviors such as pausing and jumping back and forth during the online learning process.

3. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The face recognition module is built based on FaceReader software, and the training database of FaceReader software contains facial expression data of the elderly and East Asians.

4. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The machine learning data mining method is cluster analysis, which is used to calculate the sum of clustering errors of all samples of facial emotion data of the elderly.

5. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The smooth line graph is drawn using the expression recognition detailed data exported from the FaceReader software using the Python language and the Matplotlib library.

6. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The data preprocessing unit is used to perform normalization processing and noise elimination processing on facial emotion data.

7. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The intervention strategy module includes a preset unit and a matching unit. The preset unit is used to define the richness classification of emotional behavior transfer based on the Z-score score table.

8. The emotional intervention system for elderly online learning based on emotional analysis according to claim 1 is characterized in that: The matching unit automatically identifies the emotion type of the elderly based on the preset emotion behavior transfer richness classification and accurately matches the corresponding intervention content.

Images