Analysis of the impact of the integration of the concept of public mental health into English teaching on the quality of the curriculum

With the development of the information age and social influence, the concept of public mental health appears more and more frequently in front of the public, and English teaching is also constantly reforming, so the English teaching system combined with public mental health (temporarily named mental health integrated teaching) should be put forward. In the contemporary context, the use of big data theory can better analyze the impact of mental health integration teaching.

Firstly, the sample sampling method based on statistics is used to sample the effective samples of the teaching course integrated with mental health. Then feature extraction and association rule mining are carried out to construct the analysis model of curriculum quality integrated with public mental health. The formula for sampling the sample data is as follows: (1) h(x)=∑1≤i≤xNi−α∑1≤i≤xNx1−α

N represents the sample sequence points of N mental health concepts for English teaching courses, and it also satisfies the association relationship, and the association relationship is as follows: (2)X=[s1,s2,...,sk]n=(xn,xn−r,...,xn−(m−1)r)

K and n- (m-1) R are expressed as the feature dimensions of the target sample data in the association relationship, and these dimensions and coefficients are used as the values for association rule mining to calculate the association integral of the analysis model of this course. The calculation formula is as follows: (3)CN,m(r)=2N(N−1)∑i=1N∑j=i+1NH(r−∥xi−xj∥)

Where, H represents the Heaviside function, which can put the sampling data and sample sequence obtained from the above calculation into a higher-dimensional model to obtain the final statistical data in the following way: (4)Q(a,bi)=∑i∑j[yij−(axij+bi)]2

Sort it out, and minimize the final value to get the set of rules for statistics: (5){ ∂Q∂a=−∑i∑j2[yij−(axij+bi)]xij=0 ∂Q∂bi=−∑i∑j2[yij−(axij+bi)]=0

When it ∂Q is the smallest, the characteristic quantities of the statistical data are as follows: (6){ a=∑i∑j(xij−x¯i)(yij−y¯)i∑i∑j(xij−x¯i)2 bi=yi−ax¯i

In this formula, the values represented by X and y are the scalars of the samples in the statistical data, and they are still effectively calculated. Relevant methods are used to extract the statistical data in the whole model by using big data, and the course quality analysis model integrated with public mental health is associated and solved. The process is as follows: (7)R(i,j)−H(εi−dij) (8)G(U|μk,∑k)=(2π)−d/2|∑k|−1/2*exp[−12(U−uk)T∑k−1(U−uk)]

G(U|μk,∑k) Under the solution of this association rule, a more intelligent calculation can be carried out on the course quality analysis model integrating public mental health.

The performance calculation of the model is equivalent to the analysis of English teaching effectiveness after the integration of mental health in the research object of this paper. The next step is to compute the optimal estimate of the associated feature of the fused model sample data as follows (9)Dt+1′=1−(1−λ)∑n=0∞Ωm+n+1(n+1)(b−t) (10)Lt+1′=(1−λ)∑n=0∞Ωk+n(n+1)b

where Ω is expressed as the entropy in the calculation of the correlation feature, then the calculation of the effective correlation feature is as follows: (11)xk=f{xk−1,uk−1,wk−1}

U represents the model parameters of the course quality analysis model integrated with public mental health, and w is the statistical value calculated in the model. According to the above formula, a new algorithm model for feature extraction is constructed, which is tested and controlled by the classification function. The model is as follows (12){ x(k+1)=x(k)+kx(k−τk)+w(k) z(k)=x(k)+u(k)+w(k)

According to the mining method of association rules, the sequences of X (t) and Z (t) in the model are analyzed, and the variables in the control test are calculated, and the evaluation model is obtained: (13)xij=[lgr]ij,yij=[lgCn(r)]ij

The feature points are recorded as: (14)yij=axij−bi

Finally, combined with the relevant methods of feature analysis, the random probability of a sample data appearing in a certain place is obtained. According to the spatial principle, the correlation dimensions are as follows (15)D=limr→0lgCm(r)lgr

C is the correlation integral of some data in the course quality analysis model. When the Euclidean distance is small enough, the relevant component can be calculated.

Because of the integration of the concept of public mental health, the emotional analysis of students is a very important part. By extracting the characteristics of students in the classroom to make emotional judgments, we can better analyze the mental health of students, and after understanding the concept of students’ mental health, we can make a correct evaluation of the quality of the classroom, so when optimizing the model, we choose to use the neural network in artificial intelligence as the basis, and apply it to the algorithm to build a model for analyzing students’ emotions. Two kinds of neural network models are introduced: 1)

bag of continuous words

The continuous bag-of-words model is a neural network model with a special structure of three layers. The model used in this paper is mainly used to select the target content in the article to predict, obtain the probability of the current content in the full text, and judge the concept of students'psychological status according to this probability. The calculation formula is as follows: (16)P(wt|wt−k,wt−(k−1),wt+1,...,wt+k)=P(wt|context) 2)

Jump word model

The principle of this model is almost the same as that of the continuous bag-of-words model, and it also has a three-layer structure, but it is mainly used for the selection and prediction of local information. Compared with the continuous bag-of-words model, the range is smaller, but in the same range, the accuracy of the model is higher, and the vectors are more similar. The formula is as follows: (17)P(wt|wt−k,wt−(k−1),Λ,wt+1,Λ,wt+k)=P(wt|context)

Convolutional neural network has shown its superiority in deep learning and feature extraction models. It not only plays a role in these two models, but is also widely used in other aspects. This paper will apply the convolutional neural network model to computer vision, so that the computer can accept visual input, replace manual feature extraction, so that it has the ability to automatically extract students’ emotional characteristics, and obtain students’ mental health concept from these data, so as to better reflect its impact on English classroom teaching.

There are four special structures of neural network, and the operation of each layer is different, each with its own particularity. The operation of the convolutional layer is a key calculation step, and the expression is as follows: (18)ci=f(ωTxi:i+h−1+b)

This operation can enhance the original features, keep the original statistical data unchanged, but reduce the total amount of calculation, thus reducing the final information loss rate. F is an activation function that convolves the entire step as follows: (19)c→=[c1,c2,...,cn−h+1]∈Rn−+1

This part is based on the Transformer model to improve and optimize the original integration of public mental health course quality analysis model. Its application reduces the application of the text library and improves the efficiency of emotional feature extraction. In this part, the weight of the target data should be calculated first. Before the weight calculation, the similarity between the target data and the emotional features should be calculated. The two formulas are as follows: (20)f(Q,K)=QTK (21)f(Q,K)=QTWaK

The weight data is spliced, and the calculation formula of the perceptron is as follows: (22)f(Q,K)=VaTtanh(waQ+UaKi) (23)f(Q,K)=wa[QKi]

Finally, using the softmax function for normalization, we get: (24)ai=sofymat(f(Q,Ki))=exp(f(Q,Ki))∑exp(f(Q,Ki)) (25)Attention(Q,K,V)=∑iaiVi

So far, the improvement and optimization of the course quality analysis model integrated with public mental health is over. The three optimization models can theoretically improve the feature extraction rate and reduce the data loss rate, and standardize and normalize all the structures in each part of the data to facilitate the analysis of emotional features.

Using mean to test the accuracy of the algorithm, compare the common algorithm 1 and algorithm 2 on the market with the basic algorithm and optimization algorithm in the article, and obtain the data before joining the public mental health test and the data after joining the public mental health test. Compare the four algorithms, the higher the score, the higher the accuracy, as shown in the following table 1:

Through the mean test algorithm detection accuracy in Figure 1, it can be concluded that the optimization algorithm has the highest score before and after joining the public mental health, and then the optimization algorithm is more accurate.

Figure 1.

Algorithm comparison

In order to intuitively show the change trend of UPI total score with years, the UPI total score of each year and the percentage of three types of results are drawn into a broken line chart by year.

In Figure 2, The results showed that although the mean of UPI test fluctuated, it showed a linear downward trend. The proportion of the first kind of serious psychological problems and the second kind of general psychological problems also showed a decreasing trend, but the proportion of the third kind of mental health increased gradually. This year, the mental health level of college students is increasing year by year. After analyzing the mental health of different college students, it is found that different genders, growth backgrounds and other factors have different influences on students. Among them, there are obvious differences in whether it is an only child or not. Combined with the analysis of previous research results, parents’ education level directly affects children’s mental health and development.

Figure 2.

UPI value change

Whether the mental health of the subjects in English class leads to the change of course quality (change = 1, no change = 0) is selected as the dependent variable, and the indicators with statistical significance in the univariate analysis and the four comprehensive scores of teaching quality are selected as the independent variables, and the stepwise forward method (likelihood ratio) is used to carry out logistic regression analysis. The results show that the comprehensive scores of the four common factors of curriculum quality, teacher ability, facility environment and external support are shown in the figure 3 and Table 2 below:

Figure 3.

Factor Influence Rate

Factor analysis is used to calculate the comprehensive score of teaching quality evaluation, which reduces the number of variables. Select 18 common factors with a factor satisfaction degree of not less than 0.5 as the basis for weighing. The model of the relationship between the factors and the teaching quality is established by using the stepwise regression analysis method, and the significance of the model is verified. The score intervals of the four common factors are divided into 10 equal signs, and the scores from low to high correspond to the scores of 0 to 1. The conclusion is that the factors such as accommodation environment, teacher ability, facility environment and external support will affect the quality of the course, and the teacher ability is the most important factor affecting the quality of the course.

The study uses Spearman’s correlation analysis method, with social support as the independent variable, X, active coping style as the intermediary variable, the total score of M. CCSMHS is Y, and its intermediary effect is analyzed. Data were collected from 396 college students through a questionnaire. Using SPSS statistical software on the data obtained from the exploratory factor analysis and confirmatory factor analysis, the results are as follows in Figure 4 and Figure 5:

Figure 4.

Exploratory factor analysis

Figure 5.

The influence rate of mental health on English classroom

The results showed that social support had a significant positive predictive effect on positive coping style (β = 0.4 9 1;t=2 8.6 Social support was a significant negative predictor of mental health (β < 1.1 7 6;t=8.2 Positive coping style can negatively predict mental health (α < 0.3 2 0;t=2.4 7 7 7); all three paths of the model were statistically significant (P < 0.0 5)。 Therefore, positive coping style (SSRS) played a partial mediating role in the relationship between positive coping style (CCSMHS) and mental health. Spearman correlation analysis showed that SSRS was subjective support. There was a negative correlation between the utilization dimension and the scores of each dimension of CCSMHS (P < 0.05), and there was a significant negative correlation between the SCSQ and the scores of each dimension of CCSMHS.

The English teaching quality evaluation scores of 317 subjects were collected, in which the course quality dimension included four items, namely, reasonable curriculum and schedule, teaching content focusing on theory, experimental practice meeting the needs of the course, and improving personal practical ability; the teacher ability dimension included Five items, teachers are enthusiastic about teaching, teachers can teach students in accordance with their aptitude, teacher-student interaction is smooth in the teaching process, extra-curricular learning is guided by teachers, teachers will help to develop learning objectives; the dimension of facility environment includes Three items, schools provide safe learning environment, schools provide psychological counseling and problem counseling. The students’ dormitory environment is elegant and comfortable. By comparing the total scores of the selected students with the scores of the national norm, the impact of mental health education on the quality of English classroom teaching is obtained, as shown in the following table 3 and Figure 6:

Figure 6.

Comparison of student scores with the national student norm

Under the 12 items of teaching quality evaluation, the top three items were that the teaching instruments and equipment could meet the learning needs [ (7.37 ± 2.98)], the school provided venues and conditions for scientific research activities [ (7.40 ± 1.01)], and the dormitory environment was elegant and comfortable. Mental health education plays a positive role in improving the quality of English classroom teaching, but the quality of English classroom teaching needs to be further improved due to the constraints of school teachers’ ability and other factors.

Out of 3304 persons, 37.1 were male, 62.1 were female, 54.1 were primarily urban, and 13.1 needed specialized mental health services to reduce the impact of their mental health environment. There is a significant correlation between the two variables of mental health service and mental health environment, and the CC-SMCS model has a good fitting effect. Through the analysis of the basic situation of students, it can be found that the mental health level of students in different genders and grades is compared, as shown in the figure 7 and Table 4 below:

Figure 7.

Comparison of mental health levels

It can be concluded that among the 3304 people, 37.3 were male, 62.6 were female, 54.1 lived in cities, and 13.1 needed professional mental health services to reduce the impact of the epidemic on them. The senior high school group (first grade) is better than the junior high school group and the primary school group; the difference between the types of schools is not obvious; and relatively speaking, if the economic level of students is relatively high, the level of mental health is relatively poor. Mental health service plays a positive role in these two aspects, but there are also some problems, such as teachers’ insufficient attention to mental health education, teachers’ insufficient understanding of the quality of English classroom and so on.