Study on the Construction of Informatization Platform for Civic and Political Education Management in Colleges and Universities and Its Enhancement of Students’ Educational Effectiveness

With the continuous development and maturity of information technology, digital student work management platform has become a very important part of school education and management. In particular, the construction of digital student work management platform based on ideological and political education can better play the role of ideological and political education and promote the comprehensive and healthy growth of students [1-4].

The informatization platform for the management of ideological and political education in colleges and universities has the following important significance: first, to improve management efficiency. The traditional management of student work mostly relies on manual records, document transmission, etc., which is inefficient and prone to problems such as poor information and chaotic workflow. The use of information technology platform can realize information management, improve management efficiency, and greatly save teachers’ time and energy [5-8]. Secondly, it is to promote the digitalization of education and teaching reform. The informatization platform can provide more teaching resources and information for college education, promote education and teaching reform, and promote the sharing of teaching resources and the use of quality educational resources [9-10]. Thirdly, it promotes the comprehensive quality evaluation of students. Digital student work management platform can be based on the performance and achievements of students to carry out a comprehensive evaluation, not only to evaluate students’ academic performance, but also to evaluate their comprehensive quality, to promote the overall development of students [11-14]. Ideological and political education as an important part of the educational work of colleges and universities, especially in the current social development and student growth environment, the importance of ideological and political education is more prominent. The construction of informatization platform based on ideological and political education is imperative and has irreplaceable significance [15-18].

This article explores a method of constructing an informatization platform for the management of ideological education in colleges and universities. The article firstly summarizes the general process of constructing the informationization platform for Civic and Political Management. Aiming at the shortcomings of K-means algorithm in the clustering of Civic and political information, which is easy to make the algorithm fall into the local optimum, the Canopy algorithm is used to carry out coarse clustering of the data before the K-means clustering, which improves the speed of clustering. The maximum and minimum distance algorithm is used to select as far as possible sample points for the K-means algorithm as the initial clustering center to avoid the K-means algorithm from falling into the local optimum. Use the optimized clustering algorithm to process the data of ideological education management and construct the information platform. Finally, the comparative experiment examines the effect of this paper’s platform on the improvement of the effectiveness of students’ ideological education.

2

The general process of constructing the informationization platform of ideological management

Civic education management informatization platform of colleges and universities is a kind of informatization platform based on students’ education and management information, constructing students’ portraits for accurate civic education, aiming at improving students’ education effectiveness. The general process of constructing the informationization platform for civic and political education management is as follows.

2.1

Data collection

With the development of the information society and the construction of smart campuses, students’ basic information, class attendance, competition participation, examination information, practical activities, library borrowing information, consumption information, club activities, access records, graduation design, participation in enterprise presentations, and examination for graduate school during their school years, as well as their behavioral trajectories on platforms such as class groups, circles of friends, and forums, etc. are all used as raw data, reflecting their personality traits, learning status, consumption status, employment status, interests and hobbies, and other attributes. These data reflect students’ personality characteristics, learning status, consumption status, employment status, hobbies and other attributes. The above static or dynamic multi-source and multi-dimensional data form the basic data for building student profiles.

2.2

Data analysis

The huge amount of data collected is complicated and massive, and it needs to be cleaned, screened, counted and mined by relying on big data processing technology, as well as monitoring the dynamic data and updating it in time, which is an important part of the precise civic politics of the university civic workers. Through data analysis, useless information is filtered, similar information is merged and processed, different structural data is adjusted and unified, and the hidden laws behind the data are unearthed.

2.3

Student Portraits

Labeling the analyzed and processed data, converting students’ learning, life, behavior, psychology and other data into a characteristic classification label, such as good grades, academic difficulties, family financial difficulties, work and rest patterns, etc. On this basis, using clustering analysis, correlation analysis and other means, we construct students’ individual portraits and group portraits, and label their characteristic attributes to comprehensively display students’ behavioral trajectories.

2.4

Portrait application

Combined with traditional ideological and political work ideas and experience, we analyze the constructed student portraits, accurately grasp the students’ individual needs and behavioral trajectories in learning, life, employment, social interaction, etc., and realize the dynamic monitoring of the students, on the basis of which, we accurately provide the contents of the ideological and political work and formulate the personalized education program to realize the relevance and effectiveness of the ideological and political education.

3

Data mining and clustering algorithm based on students’ civic and political information

3.1

K-means clustering algorithm

K-means is an iterative solution algorithm, where $K$ refers to the number of clustering clusters and Means is understood as the mean value of the data in each clustering family, so the K-means algorithm is also often referred to as a $K$ -means algorithm. Although K-means has been proposed for more than sixty years, it is still widely used in various fields, such as document clustering, image segmentation, feature learning, etc., due to its advantages of simplicity and efficiency [19].

For dataset $D = \{x_{1}, x_{2}, \dots, x_{m}\}$ , the K-means algorithm can divide all the samples in the dataset into $K$ clusters according to the distance between the samples in the dataset and the cluster centers. The least square error $E$ of the set of cluster centers $C = \{C_{1}, C_{2}, \dots, C_{k}\}$ in the clustering result is calculated as shown below: (1) $E = \sum_{i = 1 x = C_{i}}^{k} {‖x - μ_{i}‖}_{2}^{2}$ $$E = \mathop \sum \limits_{i = 1x = {C_i}}^k x - {\mu _i}_2^2$$ (2) $u_{i} = \frac{1}{|C_{i}|} \sum_{x = C_{i}} x$

The magnitude of the value of $E$ in Equation 1 represents the closeness between the sample points in each cluster and the clustering center. And $u_{i}$ in Equation (2) is the mean vector of $C_{i}$ . There are many ways for the K-means algorithm to calculate the distance between the samples in the dataset, among which two methods, Euclidean distance and Minkowski distance, are more commonly used.

The Euclidean distance is calculated as shown in equation (3): (3) $d (x, y) = \sqrt{\sum_{i = 1}^{m} {(x_{i} - y_{i})}^{2}}$

The Minkowski distance is calculated as shown in equation (4): (4) $d (x, y) = {(\sum_{i = 1}^{m} {|x_{i} - y_{i}|}^{p})}^{\frac{1}{p}}$

K-means algorithm for large-scale data clustering analysis, often can achieve good results, but the K-means algorithm also has some problems, such as the number of clusters and the initial clustering center need to be confirmed by humans, if these two parameters are not reasonable, it is easy to make the algorithm fall into the local optimum. Aiming at the problems of the K-means algorithm, this paper follows up with the relevant optimization.

3.2

Data processing

Various types of information management systems in colleges and universities have been operating for many years, storing a huge amount of student behavioral data, which are not in a uniform format, and there are also some missing data, which can’t be directly used for data mining. Therefore, the original data need to be preprocessed before using the improved K-means algorithm for clustering analysis to ensure that the requirements of student behavior analysis are met and the quality of data mining results is improved. The main operations are as follows:

3.2.1

Data cleansing

For the grade data in the teaching system, some students’ data are missing due to uncertainty factors such as suspension and absence from exams, so in order to ensure the integrity of the data, these data are cleaned. Combined with the student registration information, the data of suspended students are eliminated, and the grades of absent students are recorded as 0. Similarly, the data of consumption data, attendance data, library borrowing data, and access control data are carefully analyzed to make the corresponding data cleansing.

3.2.2

Data conversion

In this paper, statistical methods are used to transform student behavior data to make the data more valuable for analysis. For example, for the student one-card data, the average monthly consumption, consumption frequency, consumption peak and other indicators are counted by month as a cycle, for the library data, the average monthly borrowing times, the number of times students go in and out of the library, the library study hours and other indicators are counted, for the faculty system grade data, the average score of students in each semester, the course pass rate and other indicators are extracted to represent the semester’s students’ The overall level of students’ performance in this semester.

3.2.3

Standardized and normalized processing

There is a great deal of variability among student data indicators, which can seriously affect the results when performing cluster analysis. In order to eliminate the interactions between indicators, data need to be standardized and normalized. The normalization method is shown in equation (5): (5) $X_{i j}' = \frac{X_{i j} - X_{\min}}{X_{\max} - X_{\min}}$ where $X_{i j}'$ denotes the normalized value of the attribute, $X_{i j}$ denotes the un-normalized value, $X_{\max}$ denotes the maximum value of all samples of the attribute, and $X_{\min}$ denotes the minimum value of all samples of the attribute.

3.3

K-means algorithm based on Canopy and maximum minimum distance

3.3.1

Determination of the number of clusters for clustering

Canopy algorithm is a fast clustering technique that gives the optimal number of clusters although the accuracy is low and does not give exact cluster results.Canopy algorithm involves setting two variables $T_{1}$ and $T_{2}$ , randomly selecting the initial clustering centers, calculating Euclidean distances between the samples and the initial centers, and partitioning the initial dataset into multiple incompletely overlapping subsets based on thresholds $T_{1}$ and $T_{2}$ .

The Canopy algorithm is less accurate than the K-means algorithm, but it has a great advantage in speed and does not need to specify the number of clusters beforehand, so the Canopy algorithm can be used to carry out “coarse” clustering of the data, and then the number of clusters and the cluster center obtained by the Canopy algorithm are used as input parameters for the K-mean algorithm to complete the “fine” clustering of the data. The number of clusters and cluster centers obtained by the Canopy algorithm are then used as input parameters for the K-mean algorithm to complete the “fine” clustering of the data. However, using the cluster centers of Canopy as the initial clustering centers of the K-means algorithm may cause the K-means algorithm to fall into a local optimum [20]. Therefore, only Canopy’s algorithm is chosen to determine the number of clusters $K$ . The specific algorithm is as follows:

Input: dataset $X = \{x_{1}, x_{2}, \dots, x_{n}\}$ .

Step 1: Determine two distance thresholds $T_{1}$ and $T_{2}$ , where $T_{1}$ is greater than $T_{2}$ , by cross-validating the tuning parameters or prior knowledge.

Step 2: Randomly select a sample point $P$ from dataset $X$ as the first Canopy clustering center and add $P$ to set $C$ . Remove $P$ from $X$ .

Step 3: Randomly select a sample $Q$ from set $X$ and compute the Euclidean distance from $Q$ to each of the Canopy clustering centers in set $C$ , choosing the smallest of these distances.

Step 4: Compare $T_{1}$ with distance $d$ . If $d$ is greater than $T_{1}$ , add sample point $Q$ to the Canopy with the smallest distance from it and label it with a weak marker. If $d$ is less than $T_{1}$ , set $Q$ as the new Canopy clustering center, add sample point $Q$ to set $C$ , and remove $Q$ from data set $X$ .

Step 5: Comparison will be made with distance $d$ . If $d$ is less than $T_{2}$ , attach a strong marker to it, update the center position of all strongly marked samples to the clustering center of this Canopy, and remove sample point $Q$ from dataset $X$ . If $d$ is greater than $T_{2}$ , add $Q$ to the current Canopy set.

Step 6: Repeat steps 3 through 5 until dataset $X$ is empty.

Output: number of samples $K$ in cluster center set $C$ .

3.3.2

Selection of initial clustering centers

In K-means algorithm, the selection of initial clustering centers is random. Random selection of initial clustering algorithm falls into local optimum, and if the initial clustering centers are clustered together, it will increase the number of iterations of the algorithm and consume more time, so the selection of initial clustering centers is an important part of the improvement of K-means algorithm.

Maximum minimum distance algorithm is a trial-based clustering algorithm which can use Euclidean distance to select K clustering centers as far as possible. Using the maximum minimum distance algorithm for K-means algorithm to select as far as possible the sample points as the initial clustering center, can avoid the K-means algorithm in order to avoid falling into the local optimum and can reduce the number of algorithm iterations [21]. The specific algorithm is as follows:

Inputs: dataset $X = \{x_{1}, x_{2}, \dots, x_{n}\}$ , number of clusters $K (K \geq 3)$ .

Step 1: Select a random sample from dataset $X$ as the first cluster center $c_{1}$ .

Step 2: Using Euclidean distance, select the sample with the farthest distance $c_{1}$ as the second cluster center $c_{2}$ .

Step 3: Calculate the Euclidean distance $d_{i j}$ (where $i = 1, 2, \dots, n$ ; $j = 1, 2$ ) between each clustering center and the samples in dataset $X$ , and find the minimum of these. Select the sample point corresponding to the maximum value $d_{l}$ in $\{d_{1}, d_{2}, \dots, d_{n}\}$ and set it as the third clustering center $c_{3}$ . Set $m = 4$ , then: (6) $\{\begin{matrix} d_{i j} = ‖x_{i} - c_{j}‖ \\ d_{i} = \min [d_{i 1}, d_{i 2}] \\ d_{l} = \max \{d_{1}, d_{2}, \dots, d_{n}\} \end{matrix}$ $$\left\{ {\matrix{ {{d_{ij}} = \parallel {x_i} - {c_j}\parallel } \cr {{d_i} = \min \left[ {{d_{i1}},{d_{i2}}} \right]} \cr {{d_l} = \max \left\{ {{d_1},{d_2}, \ldots ,{d_n}} \right\}} \cr } } \right.$$

Step 4: Select the $m$ st clustering center. For $j = 1, 2, \dots, m - 1$ , calculate the distances $d_{i j}$ from the samples in dataset $X$ to all the clustering centers in set $\{c_{1}, c_{2}, \dots, c_{m - 1}\}$ , respectively, and calculate $d_{l}$ by Equation (7). select the sample corresponding to $d_{l}$ as the $m$ th clustering center $c_{m}$ : (7) $d_{l} = \max_{i} [\min [d_{i 1}, d_{i 2}, \dots, d_{i j}]]$

Step 5: If $m$ is equal to $K$ , output clustering center set $C$ ; otherwise $m = m + 1$ , repeat step 4.

Output: initial clustering center set $C = \{c_{1}, c_{2}, \dots, c_{k}\}$ .

3.4

Data mining and clustering experiments

3.4.1

Data sources

The source of experimental data in this paper is the accumulated historical record data from the management system of various campus departments made from the data in the digital campus shared library of a university, and the original data include the undergraduate students of this university from April 2022 to April 2024: library book borrowing record data, academic classroom attendance record data, library access control record data, as well as the consumption of the students, achievement data, physical exercise record data, and the wireless network access record data of the campus. The data will include: library book borrowing data, academic class attendance data, library access control data, student consumption and performance data, physical activity data, and campus wireless network access data. Since the purpose of this section is to validate the effectiveness of the clustering algorithm used, only the results of the student consumption patterns and study effort breakdowns are selected here for detailed analysis and illustration.

3.4.2

Algorithm Validation

We evaluate the results of clustering by an evaluation criterion V that reflects the similarity within classes of inter-class dissimilarity and compare it with the traditional K-means clustering algorithm, Figure 1 shows the results of the clustering evaluation criterion.

When we set the student density to 400, the clustering evaluation criterion achieves optimality. At the same time, when the student density threshold is set high, the students are divided into fewer classes, the interclass dissimilarity is small, and the intraclass aggregation is very poor, so the V value is taken to a negative value. As we adjust the student density threshold, the clustering evaluation criterion parameter rises and the clustering effect is better, but when our student density drops to a certain level, the inter-class dissimilarity is decreasing but the intra-class aggregation effect decreases significantly, resulting in a decrease in the parameter V. It shows that compared with the traditional K-means algorithm, the clustering effect of K-means algorithm based on Canopy and maximum minimum distance has a relatively obvious improvement, which is more suitable for the clustering analysis of students’ behavioral patterns.

3.4.3

Consumption information clustering results

The optimized K-means algorithm is used for clustering on Spark platform, and the results are shown in Table 1, in which “1” is used to mark that the index of the cluster is greater than or equal to the overall students’ average value of each index, and “0” is used to mark that the index of the cluster is smaller than the overall students’ average value of each index. The first type of students, the monthly average of the indicators of the first type of students.

Table 1.

Clustering results of students’ consumption rule

Cluster	Student proportion %	Monthly consumption	Monthly consumption peak	Consumption frequency	Comparison
0	10.45	529.68	618.45	238	0
1	15.22	1116.42	1249.41	452	111
2	10.69	664.29	767.08	308	1
3	7.4	616.08	971.81	213	10
4	15.25	723.73	793.10	168	100
5	18.68	865.70	886.29	346	101
6	8.52	620.52	936.66	408	11
7	13.79	794.40	938.82	205	110
AVG		741.35	895.2	292.25

The first type of students has a very low average monthly consumption, the maximum monthly consumption is not very high, and the frequency of consumption is relatively low, so it belongs to the low consumption group.

The second type of students, the average monthly consumption is the highest, the single month consumption is also very high, consumption frequency belongs to high consumption group.

The third type of students, the average monthly consumption is not high, the maximum value of single-month consumption is very low, but the frequency of consumption is relatively high, indicating that students are more active in school consumption, but the amount of consumption is limited.

Students of the fourth type have a medium level of consumption, but with high volatility, a high maximum value of consumption in a single month, and a low frequency of consumption.

The fifth type of student, whose average monthly consumption is slightly higher than the average monthly consumption of the overall student population and whose peak monthly consumption is relatively low, is closer to his or her average monthly consumption value, indicating that the student is more regular in his or her consumption at school.

The sixth type of student, with a higher average monthly consumption and a peak monthly consumption very close to the average value and lower than the average value of the overall student’s peak consumption, indicates that this type of student consumes more consistently and more frequently.

The seventh type of students, whose average monthly consumption is less than the average of the overall students, and whose peak consumption in a single month is higher but has a higher likelihood of sudden change in volatility in a particular month, indicates that their average consumption in a single month is relatively low and that they are frugal and poor students.

The eighth type of students, whose average and peak monthly consumption are both generally higher, and whose consumption frequency is less frequent, suggests that they have a higher level of consumption.

3.4.4

Clustering results of students’ civic learning information

Using this paper’s algorithm to cluster and subdivide students for their academic effort and achievement indicators, the mean values of each indicator for each type of student are shown in Table 2.

Table 2.

Clustering results of learning effort

Cluster	Student proportion %	Attendance	Achievement	Reading quantity	Library access times	Learning time	Passing rate
0	6.25	0.61	52.25	25	35.46	85	0.67
1	18.35	0.92	84.91	49	63.22	202	0.97
2	9.38	0.77	82.66	34	56.59	171	0.87
3	3.59	0.32	45.51	22	22.00	95	0.49
4	4.72	0.45	64.32	19	31.46	98	0.63
5	21.36	0.86	78.17	47	54.81	211	0.98
6	12.29	0.77	65.86	36	67.89	153	0.87
7	10.73	0.61	75.20	26	40.45	180	0.75
8	13.33	0.62	67.63	25	34.66	135	0.65

The first type of students, the weighted average grade of each subject is at the edge of passing and failing, class attendance is poor, the book reading is low and the length of study and other indicators are not very good compared with other students, so this type of students do not make enough effort, more lazy, resulting in poor grades, hovering at the edge of passing, tutors and students should strengthen supervision and strengthen the students’ self-discipline ability, in order to improve their grades.

The second type of students, the results are very good, class attendance rate is very high, the course pass rate of almost 100% pass, and other indicators are more than other students are very outstanding, so this type of students belong to the usual study is very hard and diligent “bully” type of students.

The third type of students, the grades are relatively good, but the class attendance rate is not very high, the course pass rate, the performance of other indicators are not outstanding. This type of student is intelligent but does not spend enough energy on learning, but after a sudden study to achieve better results.

The fourth type of students, the results are very poor, class attendance rate is very low, the course pass rate is not optimistic, the length of study and other indicators are the lowest, after analyzing it can be seen that this type of students often do not go to class, the course is almost always failing, which should attract the attention of teachers and parents.

The fifth type of students, the results are not particularly poor, floating in the passing area, but the class attendance rate is very low, the pass rate of the course is not high, and all the other indicators of the evaluation of the degree of effort is relatively low, it can be seen that this type of students do not love to study, but still pay more attention to the course, in the near exam by surprise review, in order to pass the exam.

The sixth type of students, the grades are in the middle to high level, class attendance attendance rate is very high, the course pass rate is very high almost no failure phenomenon occurs, and other indicators are also very high. This indicates that this type of student is particularly hard working in their studies and belongs to the practical and diligent type. However, their grades are not very good because they may not know how to study or their understanding of knowledge is limited.

The seventh type of student has a moderate level of performance and effort indicators compared to other students. For this type of student, the tutor should provide encouragement and guidance on learning methods to improve learning ability.

The eighth type of students, whose grades are moderate, and whose class attendance is low, and whose course pass rate is not high, is slightly better than the fifth type of students, but is a result of unannounced study, and is subject to some volatility.

The ninth type of students, the results are in the middle of the lower level, class attendance and course pass rate and other indicators are relatively low, indicating that this type of students in the usual degree of effort is not enough to study, although more diligent than the fifth type of students, the results are slightly better, but did not complete the task of learning to be accomplished.

Students who study hard and get good grades account for 18.35% of the students, and most of them study hard. Only a very small number (3.59%) of the students did not work hard enough and got poor grades, and a small number (4.72%) of the students passed but did not work hard enough.

According to the above evaluation criteria and the analysis of student clustering results compared with the real situation, it shows that the K-means algorithm based on Canopy and maximum minimum distance used in this paper for student behavior analysis is reasonable and reliable.

4

The Construction and Effectiveness of the Informationization Management Platform of Civic and Political Education in Colleges and Universities

4.1

Construction of Informatization Management Platform for Civic and Political Education in Colleges and Universities

4.1.1

Needs analysis

Higher education information management system is to obtain valuable information through a large amount of data accumulated on campus, to integrate this information, and to assist in the management of students in a variety of applications. The information management platform is based on students’ basic information, campus card consumption records, access control records, academic performance records and other multi-dimensional data, through the processing of big data technology, including secretary cleaning, desensitization, index calculation, feature analysis, etc., for different applications, combined with data mining technology, data classification, clustering, etc., to generate the determination of family economic difficulties, student profiles and fine management supervision of the page to help schools with information management.

4.1.2

Architectural design

The whole management platform is divided into four layers, which are data source, data processing layer, algorithm layer and application layer. The construction of the informationization platform for the management of ideological education in colleges and universities is shown in Figure 2. The data source layer is the foundation of the whole platform, which integrates heterogeneous data in the campus, effectively avoids the “information silo” of each data platform, and lays the foundation for the establishment of applications such as student behavior analysis and portrait construction. The data processing layer converts and pre-processes the data in the platform, and does feature construction and feature extraction for different applications. The portrait labeling system is constructed in the student portrait and the required indicators are calculated. The algorithm layer is based on clarifying the objectives of each module of the student information management platform, applying data mining technology to analyze the determination of family economic difficulties, the construction of the labeling system, the construction of student portraits and the management of early warning supervision. The application layer visualizes the analysis results of the third layer, develops the student information management platform, and visually displays the platform page to achieve the intelligence of actual campus management.

4.2

Experiments to improve the educational effectiveness of students

4.2.1

Analysis of learning input results

Two classes were randomly selected as the experimental class (N=50) and the control class (N=50) in the school, and the informatization platform designed in this paper was introduced in the teaching management of the experimental class, and the pre and post-test scores were evaluated by questionnaires during the experimental process, and the comparison of scores was used to test the enhancement effect of this paper’s platform on the educational effectiveness of the students.

In this study, 100 valid questionnaires were collected, and the results of the 100 post-test questionnaires were analyzed with the pre-test results of the action research by a paired-sample t-test, and the results are shown in Table 3.

Table 3.

The paired t-test results of the study involvement

Dimensions	Test sequence	N	Mean	Sd	t	Sig.
Study motivation	Pretest	100	4.663	0.902	-2.899	0.009
Study motivation	Posttest	100	5.008	1.080	-2.899	0.009
Study energy	Pretest	100	4.057	0.957	-3.602	0.000
Study energy	Posttest	100	4.501	1.062	-3.602	0.000
Study concentration	Pretest	100	4.140	1.043	-3.164	0.001
Study concentration	Posttest	100	4.576	1.244	-3.164	0.001
Overall	Pretest	100	4.285	0.896	-3.697	0.005
Overall	Posttest	100	4.692	1.044	-3.697	0.005

As shown in Table 3, the t-value of motivation dimension is -2.899, p<0.01, which indicates that there is a significant difference between the results in motivation dimension and the post-test scores are higher than the pre-test scores. In addition, the (t=-3.602, p<0.01), concentration dimension (t=-3.164, p<0.01) and overall dimension (t=-3.697, p<0.01) of the energy dimension showed statistical significance, indicating that the post-test scores were significantly higher than the pre-test scores in all dimensions of learning engagement.

As a result of the action research in the posttest, the mean values of students’ motivation, energy for learning, concentration, and overall dimensions in terms of commitment to learning increased by 0.345, 0.444, 0.436, and 0.407, respectively, compared to the pre-test. The motivation dimension had the highest mean level and all three dimensions increased compared to the pre-test, and the overall mean of the questionnaire was significantly higher. In terms of standard deviation, the standard deviation of the energy dimension of the posttest was the smallest in terms of commitment to learning, indicating that the students had the least variation in their levels of energy.

In summary, in terms of learning engagement, the motivation, energy and concentration dimensions of students’ engagement were significantly improved, as well as the overall learning engagement. It indicates that the education management informationization platform in this paper has positively affected students’ learning engagement in general.

4.2.2

Analysis of the results of the course’s political thinking

The 100 post-test questionnaire results of the first level dimensions and overall and pre-test results were analyzed by paired samples t-test, as shown in Table 4. The t-values of cultural literacy, information awareness, and overall dimensions were: were -5.611, -4.082, and -4.1, respectively, with p-values less than 0.01, showing statistical significance. And the t-statistic of value concept is -1.667, p=0.084>0.05, failing to pass the hypothesis test of significance at the 5% significance level, thus rejecting the idea that the posttest scores of the value concept dimension are significantly higher than the pretest scores.

Table 4.

The paired t-test results of each dimensions of political education

Dimension	Test sequence	N	Mean	Sd	t	Sig.
Literacy	Pretest	100	3.477	0.470	-5.611	0.000
Literacy	Posttest	100	3.822	0.519	-5.611	0.000
Values	Pretest	100	3.886	0.478	-1.667	0.084
Values	Posttest	100	3.980	0.557	-1.667	0.084
Information cognition	Pretest	100	3.560	0.470	-4.082	0.000
Information cognition	Posttest	100	3.816	0.598	-4.082	0.000
Overall	Pretest	100	3.665	0.424	-4.1	0.000
Overall	Posttest	100	3.874	0.476	-4.1	0.000

It can be seen that in terms of course ideology, the mean values of students’ cultural literacy, value concept, information awareness and overall increased by 0.345, 0.094, 0.256 and 0.209 respectively compared to the pre-test, of which all were statistically significant except for the value concept. Students’ recognition of value concepts remained the highest and the means of all three dimensions and the questionnaire as a whole increased compared to the pre-test. The standard deviation of cultural literacy was the smallest, indicating that students’ levels of cultural literacy varied the least.

The results of the paired-sample t-test analysis of the secondary dimensions of the 100 post-test questionnaire results with the pre-test results are shown in Table 5.

Table 5.

The paired t-test results of secondary dimensions of political education

Dimension	Test sequence	N	Mean	Sd	t	Sig.
Scientific spirit	Pretest	100	3.956	0.723	-1.544	0.134
Scientific spirit	Posttest	100	4.086	0.709	-1.544	0.134
Rational thinking	Pretest	100	3.180	0.563	-7.638	0.000
Rational thinking	Posttest	100	3.744	0.632	-7.638	0.000
Digital innovation consciousness	Pretest	100	3.298	0.579	-4.733	0.000
Digital innovation consciousness	Posttest	100	3.664	0.637	-4.733	0.000
Goodwill	Pretest	100	3.642	0.564	-3.799	0.000
Goodwill	Posttest	100	3.919	0.644	-3.799	0.000
Social responsibility	Pretest	100	4.027	0.642	-0.652	0.501
Social responsibility	Posttest	100	4.060	0.665	-0.652	0.501
Patriotic feeling	Pretest	100	3.911	0.706	-2.982	0.002
Patriotic feeling	Posttest	100	4.165	0.739	-2.982	0.002
Ideal belief	Pretest	100	3.989	0.621	2.058	0.04
Ideal belief	Posttest	100	3.857	0.649	2.058	0.04
Information awareness	Pretest	100	3.817	0.696	-1.042	0.319
Information awareness	Posttest	100	3.905	0.782	-1.042	0.319
Information thinking	Pretest	100	3.116	0.746	-4.899	0.000
Information thinking	Posttest	100	3.636	0.965	-4.899	0.000
Information ethics and security	Pretest	100	3.731	0.606	-2.486	0.015
Information ethics and security	Posttest	100	3.942	0.649	-2.486	0.015

The t-value and p-value of the six dimensions of rational thinking, digital innovation awareness, integrity and friendliness, national sentiment, information thinking, and information ethics and security are less than 0 and 0.05, respectively, indicating that the post-test scores of the above dimensions are significantly higher than the pre-test scores.

The p-value of the ideal belief dimension is less than 0.05, but the t-value is positive, indicating that there is a significant difference in the results of this dimension, but the post-test scores are lower than the pre-test scores.

The posttest scores for the dimensions of scientific spirit, social responsibility, and information awareness are higher than the pretest scores, but p > 0.05, and there is no significant difference between the pre and posttest results.

In terms of the mean, the more obvious enhancements are in the dimensions of rational thinking, digital innovation awareness, honesty and friendliness, national sentiment and information thinking, all of which are higher than the mean of the pre-test action research by more than 0.27, and except for the dimension of ideals and beliefs, the mean of all other dimensions is slightly higher compared to the pre-test. In terms of standard deviation, the standard deviation of social responsibility is the smallest, indicating that students’ level of social responsibility varies the least.

5

Conclusion

This study mainly improves the traditional K-means by Canopy counting and Maximum Minimum Distance algorithm, realizes the construction of students’ Civics learning portrait, and designs the Civics education informatization management platform on this basis.

1)

Clustering experiments are carried out using the data in the shared library of a university’s digital campus, and it is found that the clustering evaluation criterion achieves the optimum when the student density is set to 400. At this time, the algorithm accurately clusters student consumption and learning information into 8 and 9 classes respectively, reflecting the importance of optimized algorithms in the process of Civic Education informatization management.

2)

After the platform was put into use, the mean values of learning motivation, learning energy, concentration and overall dimensions of students in the experimental group increased by 0.345, 0.444, 0.436 and 0.407 respectively compared with the pre-test, and the difference was significant. The means of cultural literacy, information awareness and overall dimensions were significantly increased by 0.345, 0.256 and 0.209 respectively over the pre-test.

The above experimental results indicate that the clustering optimization algorithm proposed in this paper can effectively mine the information of civic management and improve the effectiveness of students’ civic education.

Sprache:: Englisch

Zeitrahmen der Veröffentlichung:: 1 Hefte pro Jahr
Fachgebiete der Zeitschrift:: Biologie, Biologie, andere, Mathematik, Angewandte Mathematik, Mathematik, Allgemeines, Physik, Physik, andere

Zeitschrift RSS Feed

Study on the Construction of Informatization Platform for Civic and Political Education Management in Colleges and Universities and Its Enhancement of Students’ Educational Effectiveness

Yi Chen

Wei Shan

Online veröffentlicht: 05. Juni 2025

Eingereicht: 29. Dez. 2024

Akzeptiert: 18. Apr. 2025

DOI: https://doi.org/10.2478/amns-2025-0977

SchlüsselwörterK-means, Canopy, Maximum Minimum Distance Algorithm, Civic and Political Education Management Platform, Educational Effectiveness

© 2025 Yi Chen and Wei Shan published by Sciendo

This work is licensed under the Creative Commons Attribution 4.0 International License.

Schlüsselwörter
K-means, Canopy, Maximum Minimum Distance Algorithm, Civic and Political Education Management Platform, Educational Effectiveness