Strategies and Practical Exploration of Digital Transformation of Party Building Work in Higher Education Institutions

Carrying out high-quality party building work is not only the basic requirement of adhering to the overall leadership of the Party over higher vocational colleges and universities, but also an urgent task facing the strengthening and improvement of party building work in higher vocational colleges and universities in the new era [1–2]. As an important part of national higher education, higher vocational colleges and universities have had more long-term development under the premise of continuous reform and deepening of China’s higher education system, and under the correct guidance of the Party and the State’s policy of active encouragement, strong support, correct guidance, and management in accordance with the law [3–6]. With the arrival of the digital era, the education concept, education system, education mode, etc. of higher vocational colleges and universities are facing an all-round adjustment and intelligent transformation, and the cognitive schema, way of thinking, and behavioral habits of teachers and students have been changed dramatically, which unavoidably has a significant impact on the content, mode and way of party building work in higher vocational colleges and universities [7–10]. Relying on digital technology to empower party building in higher vocational colleges and universities has become an inevitable trend for the high-quality development of party building in higher vocational colleges and universities in the new era. Driven by the digitalization of education, the Internet is gradually becoming the driving force for the innovation of party building in higher vocational colleges and universities, especially playing a pivotal role in extending the tentacles of party building, expanding the carriers of party building, and stimulating party building activities [11–14]. Vocational colleges bear the political responsibility of “educating people for the party and the country”, and should actively follow the law of digital development of party building, and grasp the new concepts, new methods, and new means of deep integration of party building in higher vocational colleges with new digital technologies represented by artificial intelligence, blockchain, big data, and metaverse, which will help comprehensively improve the efficiency of party building work in higher vocational colleges [15–18].

The article firstly discusses the improvement path of party building in higher vocational colleges under the background of digital transformation, and constructs an evaluation index system of 13 secondary indicators and 39 tertiary indicators from the evaluation of the quality of students’ party building work and the performance evaluation of party building leaders at the grass-roots level of colleges and universities. Then, a logistic regression model is constructed to predict the effect of the digital transformation strategy of party building in higher vocational colleges proposed in this paper. The implementation effect is empirically analyzed using the logistic regression model constructed in this paper. Finally, the logistic regression model is used to assess the contribution of the indicators and clarify the scientificity and effectiveness of the implementation of the strategy.

2.

Research on Digital Transformation Strategy of Party Building Work in Higher Vocational Colleges and Universities

2.1

The Path of Enhancing Party Building Work in Higher Vocational Colleges and Universities

2.1.1

Innovative modes of party building work

Under the background of education digitization, the party building work of higher vocational colleges and universities should actively draw on the advanced party building concepts, management modes, and technological means of other schools to build a new mode of party building work that is suitable for the digital environment of the university, so as to enhance the relevance and effectiveness of party building work. For example, by innovating the form of combining online and offline party building activities, higher vocational colleges and universities make full use of digital tools and platforms such as online meetings and digital learning platforms to break the limitations of time and space, and expand and improve the coverage and participation of party building activities.

2.1.2

Use of technology to enhance organizational management effectiveness

Higher vocational colleges and universities can establish a digital management system to incorporate the contents of party building work into the systematic management, realize the centralized management and real-time update of party members’ information, strengthen the monitoring of party members’ participation in party building work and obtain timely feedback. In addition, the party organization of higher vocational colleges and universities can also develop and utilize mobile applications to facilitate party members to access party building information and enroll in party activities at any time and any place [19]. Various advanced technical means can provide accurate services for the party building work of higher vocational colleges and universities, so as to improve the efficiency and service quality of the party organization, simplify the management process, optimize the work link, and enhance the management effectiveness.

2.1.3

Strengthening party training and education

1)

Developing digital literacy training programs

Digital literacy includes not only the mastery and application of digital technology, but also aspects such as network security awareness and the ability to identify and utilize network information. In order to enhance the digital literacy of party members in higher vocational colleges and universities, the party organization should design a systematic and comprehensive digital literacy training plan and training content based on the basic literacy of party members and combined with the actual situation. This training program should not only teach digital technology and knowledge to party members, but also cultivate their information literacy and innovation ability.

2)

Utilizing online learning platforms to promote party members’ continuing education

The use of online learning platforms to promote the continuing education of party members is a crucial step towards improving the comprehensive quality of party members in the context of education digitization. The online learning platform is characterized by flexibility, convenience and interactivity, which allows party members to study anytime and anywhere, and independently choose the learning content and learning progress, thus improving the party members’ learning efficiency and learning quality. Therefore, higher vocational colleges and universities should establish their own online learning platform to provide party members with rich learning resources and useful learning tools to meet the learning needs of party members at different levels and in different fields. This learning platform can not only offer online courses on party history knowledge, party building theory, laws and regulations, but also invite experts and scholars to hold online lectures and online seminars to promote learning exchanges and thought collisions among party members, continuously improve the ideological and political quality of party members, so as to enhance the professional ability of party members, and inject new vitality into the party building work of the institutions of higher vocational education.

2.1.4

Use of social media platforms to promote party building activities

Higher vocational colleges and universities can regularly release party building activities, party theory learning, and outstanding party members’ deeds through social media channels such as WeChat public number, microblogs, QQ groups, etc., to guide party members to pay attention to and participate in them. The social media platform can facilitate the instant dissemination of information, bring party members closer to each other, and enhance the cohesion and emotional identity of the party organization. In addition, with the help of social media, higher vocational colleges and universities can expand the influence of party building activities, thus attracting more teachers, students, and the public to participate in them.

2.1.5

Strengthening data security and privacy protection

1)

Establishing data management and usage norms

In the context of education digitization, higher vocational colleges and universities should establish perfect data management and use norms to ensure the security and compliance of sensitive data. To this end, higher vocational colleges and universities should clarify the process of data collection, storage, and use, and standardize related operations. Establish a backup and recovery mechanism to prevent and respond to the risk of data loss or damage in a timely manner, and ensure comprehensive data security.

2)

Strengthen data encryption and authority control

In order to guarantee data security, higher vocational colleges and universities can adopt data encryption technology to encrypt the storage and transmission of important data, preventing the data from being illegally accessed or tampered with in the transmission and storage process. At the same time, higher vocational colleges and universities should also establish a strict authority control mechanism, assigning different access rights according to the user’s identity and role to ensure that the data can only be accessed and used under legal authorization, so as to effectively guard against the risk of data leakage and misuse.

3)

Increase the protection of personal privacy

In the digital environment, higher vocational institutions should formulate a clear privacy protection policy to protect the privacy of teachers and students and ensure the security of personal information. This privacy policy should clearly specify the purpose, scope, and manner of collecting personal information. The use, storage, and protection measures for personal information. The rights and responsibilities of both parties in the collection of information, etc. In addition, higher vocational colleges and universities should also establish a perfect privacy protection management system to strengthen the supervision and protection of personal information, so as to reduce the risk of personal privacy being leaked.

4)

Enhance the information security awareness of teachers and students

Higher vocational colleges and universities should strengthen information security awareness education and training for teachers and students to increase their attention and vigilance towards information security. Higher vocational colleges and universities can popularize information security knowledge and skills among teachers and students by carrying out information security publicity activities and organizing information security training courses, so as to strengthen their awareness of information security risks and at the same time enhance their ability to cope with the risks.

2.2

The Construction of Performance Evaluation System for Student Party Building in Higher Education Institutions

As a technical tool to objectively assess and achieve the optimization goal, the comprehensive evaluation system should contain three basic elements: indicators, scoring criteria and indicator contribution.

The selection of evaluation indicators follows the principles of scientificity, systematicity, representativeness and operability, and strives to achieve the goals of wide coverage, representativeness and operability of the selected indicators. In addition, considering that the student evaluation perspective has an important revelatory role in the performance evaluation of student party building work in higher vocational colleges, the system is divided into three aspects, including the evaluation of the quality of student party building work, the evaluation of the performance of the person in charge of grassroots party building in higher vocational colleges and the demographic characteristics of the interviewees, and the definition of the subordinate indicators in a targeted manner [20]. The dimensions and structure of the indicators used to evaluate the performance of party building work of higher vocational students are shown in Figure 1.

According to Figure 1, this system aims to objectively evaluate the performance of student party buildings in higher vocational colleges and universities, which is the target level. The first-level indicators under the target layer should reflect the evaluation dimensions, and this system deconstructs the target concept of evaluation from three dimensions: evaluation of the quality of student party building work, evaluation of the performance of the person in charge of grassroots party building in higher vocational colleges and demographic characteristics of the interviewees. At the second level, the index is the evaluation connotation level, focusing on reflecting the structure of the evaluation objectives, continuing to refine the quality evaluation of student party building work into student organization construction, system construction, party member development and education construction, publicity and demonstration and social service innovation, and refining the performance evaluation of the person in charge of grassroots party building in higher vocational colleges into five aspects: “morality”, “ability”, “performance”, “diligence” and “integrity”, and at the same time paying attention to the age, education level and political characteristics of the respondent population. The third-level indicators, i.e., the specific assessment indicators, belong to the decomposition of the connotation level of the evaluation, and continue to be deconstructed into 39 specific indicators, of which 16 are evaluations of the quality of student party building work, 19 are evaluations of the performance of the person in charge of grass-roots party building in higher vocational colleges and universities, and 4 are demographic characteristics of the interviewees. Through this process of hierarchical analysis, the students’ perspective is incorporated into the evaluation system, which also reflects the representativeness and scientificity of the selection of indicators.

2.3

Logistic regression analysis model

2.3.1

Logistic regression model

Logistic regression model predicts whether the strategy implementation effect is good or not through various characteristic data. The independent variable in the model is the data of the characteristics of the digital transformation strategy of party building work in higher vocational colleges and universities itself, and the dependent variable of the model is the effect of strategy implementation p value, p has a range of (0,1). When p ≥ 0.5, we consider the effect of the strategy implementation to be good, and when p < 0.5, we consider the effect of the implementation of the strategy to be bad [21].

We define odds (odds) as the ratio of the probability of an event occurring to not occurring, i.e.: 1 $o d d s = \frac{p}{1 - p}$

The log odds ( log odds or logit(p)) can be obtained by taking the logarithm of the chance of occurrence: 2 $\log i t (p) = \ln \frac{p}{1 - p}$

Logit function has a domain of definition of (0,1) and a domain of value of (–∞, +∞). Therefore, the output value of logit function can be expressed as a combination of characteristic variables multiplied by their respective weights. That is: 3 $\log i t (p (y = 1 | x)) = w^{T} x$

In the above equation w^T denotes the regression coefficient, i.e., the corresponding weight of the characteristic variable, x is the independent variable, i.e., the characteristic variable of the strategy, and p (y = 1|x) is the conditional probability distribution, which denotes the probability that the sample is classified as 1 when the input is x, i.e., the probability that the strategy is considered to be implemented ineffectively when the input is the value of the characteristic variable. And we want to get the implementation effect probability of the strategy when we judge whether the strategy is implemented well or not by transforming the value domain to (0,1) and the definition domain to (–∞,+∞). So we take the inverse function of the logit function: 4 $g (x) = p (y = 1 | x) = \frac{1}{1 + e^{- w^{T} x}}$

The above equation, i.e., the logistic function, is a sigmoid function, and the specific image of the sigmoid function is shown in Figure 2 diagram.

We bring in the strategy attributes and set the inter-value to 0.5, which is greater than the threshold value to be judged as default. Logistic regression model is relatively clear and direct, with strong interpretability, and is still the most widely used model in the actual scenarios, which is representative, and is often used as a benchmark for the prediction model.

2.3.2

Standardization of data

Directly substituting the original values of each variable when training the model will make the indicators with higher levels of values more useful and reduce the actual importance of indicators with lower levels of values. Therefore, it is necessary to eliminate the influence of the scale and the volatility of the variables themselves through standardization.

At present, there are various standardization methods, including the following: 1)

Min-Max standardization, through the linear mapping of the data projected onto the [0,1] interval. The conversion formula is: 5 $x^{*} = \frac{x - \min}{\max - \min}$

The Z-Score is standardized by subtracting the mean of the data point and dividing by its standard deviation. The transformation formula is: 6 $x^{*} = \frac{x - \bar{x}}{σ}$

2)

The proportional method, which mainly transforms for sequences in which all data are positive, is transformed by the formula: 7 $y_{i} = \frac{x_{i}}{\sum_{i = 1}^{n} x_{i}}$

Then the new sequence is y₁, y₂,……, y_n ∈ [0,1] and dimensionless and has $\sum_{i = 1}^{n} y_{i} = 1$ .

In this paper, Min-Max normalization method is used for processing. After the standardization of the data, Logistic regression algorithm will be used for model building, based on the learning and training of the filtered features, to evaluate the classification performance of different models and the effect of the guaranteed network metrics.

2.4

Assessment of the contribution of indicators

By summing up the values and equal-weighted attributes of the aforementioned secondary indicators, the performance scores of the student party building work of specific higher vocational colleges can be obtained, as well as their performance rankings in the group of higher vocational colleges. This system is characterized by the fact that each indicator is attribute rather than continuous, and equal weight rather than unequal discretionary, so it is of great practical significance to continue to assess the contribution of the indicators, and the results can also provide effective information for the improvement and further promotion of the Party building work of the students in higher vocational colleges and universities. The assessment method of indicator contribution involved in this system is based on logistic regression in measurement, which is a multivariate analysis method to study the relationship between the explanatory variables and the explanatory variables in the form of dichotomous or multicategorical variables, and belongs to probabilistic nonlinear regression in essence. Based on the specific conditions applied in this system, two-classified ungrouped unconditional Logistic regression is mainly adopted here. The principles of the regression model’s formulation, testing, and estimation are specified as follows:

The explanatory variable y takes values in the range of 0 and 1, and the occurrence of the event is noted as y = 1, otherwise 0. Let the explanatory variables x = (x₁,⋯,x_k)^T, n group of observations is noted as (x_i1,⋯,x_ik,y_i), where i = 1,2,⋯,n. Is noted as X_i = (1,x_i1,⋯,x_ik)^T, x_i0 = 1, then the logistic regression model of y_i with x_i1,⋯,x_ik is: 8 $\begin{array}{l} E (y_{i}) & = π_{i} = f (β_{0} + β_{1} x_{i 1} + \dots + β_{k} x_{i d}) \\ = \frac{e^{β 0 + β 1 x_{i} 1 + \dots + β h_{i k}}}{1 + e^{β 0 + β 1 x_{i} 1 + \dots + β k x_{i k}}} = \frac{e^{β^{T} X_{i}}}{1 + e^{β^{T} X_{i}}} \end{array}$

Clearly, y_i. is a 0-1 distribution with mean π_i and a distribution law of the form: $f (y_{i}) = π_{i}^{y_{i}} {(1 - π_{i})}^{1 - y_{i}}$ . The likelihood function of y₁,⋯,y_n is $L = \prod_{π_{i}^{y_{i}}}^{n} {(1 - π_{i})}^{1 - y_{i}}$ , and its log-likelihood function is of the form: 9 $\begin{array}{l} \ln L & = \sum_{i = 1}^{n} [y_{i} \ln π_{i} + (1 - y_{i}) \ln (1 - π_{i})] \\ = \sum_{i = 1}^{n} [y_{i} \ln \frac{π_{i}}{1 - π_{i}} + \ln (1 - π_{i})] \end{array}$

Substituting $π_{i} = \frac{e^{β 0 + β 1 x_{i} 1 + \dots + β k x_{i k}}}{1 + e^{β 0 + β 1 x_{i} + \dots + β x_{i k}}}$ into the above equation yields: 10 $\begin{array}{l} \ln L & = \sum_{i = 1}^{n} [y_{i} (β_{0} + β_{1} x_{11} + \dots + β_{k} x_{i k}) - \ln (1 + e^{β 0 + β 1 x_{i} 1 + \dots + β k x_{i k}})] \\ = \sum_{i = 1}^{n} [y_{i} β^{T} X_{i} - \ln (1 + e^{β^{T} X_{i}})] \end{array}$

Noting LL(β) = ln L (β), the estimate β = (β₀,β₁,⋯,β_k)^T of β = (β₀,β₁,⋯,β_k)^T is chosen so as to make LL(β) extremely large, this is the great likelihood estimation of the Logistic regression model, and the solution of this process requires the use of Newton’s iterative method: constructing the score function F_g (β) = ∂LL(β)/∂β_z, where g = 0,1,⋯,k, for a total of k +1 systems of nonlinear equations, making it equal to 0 and solving for β, where: 11 $F_{g} (β) = \sum_{i = 0}^{n} (y_{i} x_{i g} - \frac{x_{i g} e^{β^{T}_{x_{i}}}}{1 + e^{β^{T}_{x_{i}}}})$

Construct the information matrix I_gh (β) = –∂² LL(β)/∂g∂h, where h = 0,1,⋯,k, i.e., the negative matrix of the LL(β) second-order derivative matrix, where: 12 $I_{g h} (β) = \sum_{i = 0}^{n} \frac{x_{i g} x_{i h} e^{β T_{x i}}}{{(1 + e^{β^{T}_{x i}})}^{2}}$

I_gh(β) = I_hg (β) from the above equation, so I(β) is a symmetric matrix.

In the actual assessment of the contribution of indicators, we can make the student evaluation of the grassroots party building work in higher vocational colleges and universities y as a binary attribute variable, which takes the value of 1 if a student’s evaluation of the party building work in higher vocational colleges and universities and the work of the person in charge of the party building in a five-point Likert scale is more than the weighted average score, and takes the value of 0 otherwise.A linear probabilistic model to portray the student evaluation of the party building work performance is constructed as follows: 13 $P = \frac{\exp (β_{0} + β_{1} x_{1} + \dots + β_{m} x_{m})}{1 + \exp (β_{0} + β_{1} x_{1} + \dots + β_{m} x_{m})}$

Where P represents the results of students’ evaluation of the performance of party building in higher vocational institutions, taking only two discrete values of 0 and 1. x. Denotes a series of secondary indicators within the system. β. Denotes the natural logarithmic value of the ratio of high party building performance to low party building performance when the independent variables are all zero. β₁,⋯,β_m represents the partial regression coefficient, which means the degree of contribution of the factors x. to P.

3

Empirical analysis

3.1

Logistic regression analysis

3.1.1

Parallel lines check

Logistic models need to satisfy the proportional dominance assumption, that is, regardless of the location of the split point of the dependent variable, the effect of each independent variable in the model on the dependent variable is unchanged, and the regression coefficient of the independent variable on the dependent variable is independent of the split point. The ordered multicategorical Logistic model’s regression principle involves splitting multiple classifications of the dependent variable into multiple binary Logistic regressions sequentially. The ordered multicategorical Logistic regression model must test the proportionality advantage assumption (also known as the parallel lines test), and if it does not pass the parallel lines test, the use of an unordered multicategorical Logistic regression model needs to be considered. The parallel lines test is shown in Table 1. From the table, the log likelihood value of the independent variable is 0, which does not pass the parallel lines test. Therefore, the unordered multicategorical Logistic regression model is used.

Table 1.

Parallel test

	Logarithmic likelihood	Cabs	Freedom	Significance
Original hypothesis	0.0000	/	/	/
Routines	0.0000	0.0000	25.0000	1.0000

3.1.2

Modeling

If a variable has more than two categories, but these categories cannot be prioritized in order, these variables are unordered multicategorical variables. In this paper, using SPSS software, the unordered multicategorical Logistic model is used to estimate the indicators, and the results of the model fitting information are obtained, and the model fitting information is shown in Table 2. From the table it can be seen that the log-likelihood value of the model with the addition of the independent variable is 73.6511, and the log-natural value of the model with only constant terms is 145.6233, and the model with the addition of the independent variable has a better fit than the model with only constant terms. The significance level below 0.001 indicates that the inclusion of independent variable x is statistically significant.

Table 2.

Model fitting information

Model	Logarithmic likelihood	Cabs	Freedom	Significance
Intercept	145.6233	/	/	/
In the end	73.6511	75.0035	22.0000	0.0000

The results of the likelihood ratio test of the unordered Logistic regression model are shown in Table 3. As can be seen from the table, the factors of student organization construction, system construction and party member development and education can explain 48.7% of the variation in the respondents’ satisfaction with the digital transformation policy of party building in higher vocational colleges and universities. For example, the regression coefficient value of the factor of student organization construction is 1.539 and shows significance at the 0.05 level (z=4.112, p=0.038<0.05), implying that student organization construction has a significant positive influence relationship on the respondents’ satisfaction with the digital transformation policy of party building in higher vocational colleges and universities, as well as the value of Exp(B) is 4.598, implying that The probability of change (increase) in respondents’ satisfaction with the policy of digital transformation of party building in higher education institutions is 4.598 times when the interval of student organization building increases by one unit.

Table 3.

The disorder logistic regression model is comparable to the test results

	Regression Coefficient	Standard Error	Z Value	Exp(B)	Significance	OR Value 95% CI(LL)	OR Value 95% CI(UL)
Student Organization Construction	1.539	0.737	4.112	4.598	0.083	0.996	19.829
Institutional Construction	1.136	0.449	5.971	0.317	0.014	0.089	0.833
Member Development Education	1.316	2.331	6.798	0.011	0.000	0.042	0.036
Demonstration	-0.382	0.847	0.224	0.685	0.673	0.188	3.299
Social Service Innovation	-0.375	1.306	0.076	0.682	0.773	0.09	8.291
“DE”	0.928	0.729	1.39	2.487	0.229	0.546	11.216
“Energy”	-1.396	0.466	10.648	0.242	0.024	0.155	0.607
Gpa	-0.072	1.273	0.006	0.921	0.955	0.065	11.977
Attendance	1.149	1.687	13.904	4.762E-9	0.012	0.032	0.011
Incorruptibility	18.644	3.013	37.708	12.434	0.024	32.546	4.778E10
Age	1.553	0.77	4.135	4.538	0.101	0.93	19.777
Education Degree	1.108	0.43	5.943	0.277	0.008	0.123	0.809
Political Appearance	1.283	2.333	6.831	0.029	0.054	0.097	0.04
Grade	-0.409	0.861	0.303	0.641	0.612	0.169	3.356
Mcfadden R²:0.487
Cox Andsnell R²:0.615
Nagelkerke R²:0.736

3.2

Predictive analysis of the model

In order to test the predictive accuracy and stability of the model, the data of party building work of a higher vocational college in 2023 is used as the data sample of the test group, which is substituted into the principal component Logistic model constructed above, and the calculation can get the effect of strategy implementation p. Generally, the cut-off value of 0.500 is used as the cut-off value of the test, and the rule of determination is: when p>0.500, that is, the probability of the strategy implementation is ineffective greater than 50%. When p ≤ 0.500, that is, the poor effect of strategy implementation is less than 50%. The model predictive ability test is shown in Table 4. For the selected test group samples, it can be seen from the table that the Logistic model has a prediction accuracy of 76.91% for strategy implementation with Z value ≤ 2.575. The prediction accuracy of the Logistic model for the implementation of strategies with Z-value > 2.575 is 85.22%.The prediction accuracy of the Logistic model for the overall sample group samples is 84.60%. It can be seen that the Logistic model has good validity and stability in predicting the implementation of strategies for the digital transformation of party building in higher vocational colleges and universities.

Table 4.

Model prediction test

Observed		Predicted
		Z Value Virtual Variable		Percentage Correction
		0	1	Percentage Correction
Z Value Virtual Variable	0(Z<2.575)	115	26	76.91
Z Value Virtual Variable	1(Z>2.575)	55	237	85.22
Total Percentage				84.6
The Cut Value Is 0.500

3.3

Comparative analysis of the strength of interpretation of indicators

In this section, data related to party building in a higher education institution for the year 2023 is used for the study. The SHAP of the logistic regression model before and after the implementation of the strategy is shown in Figures 3 and 4. In Figures 3 and 4, the vertical coordinates indicate each input feature, and the horizontal coordinates indicate the Shapley values corresponding to each feature. As can be seen from Figure 3, in the pre-strategy implementation logistic regression model, the top four indicators that explain the model output results most strongly are political appearance, education, “performance”, and “ability”. As can be seen in Figure 4, in the logistic regression model after the implementation of the strategy, the top four indicators that contribute the most to the model output are political appearance, education level, “integrity” and “diligence”.

In addition, the Shapley values of each sample feature can be combined into the interpretation of the model output for the whole by averaging and then taking the absolute value, which is known as the SHAP feature importance, and based on the different degrees of feature attribution, the SHAP feature importance can replace the original feature importance method. According to the SHAP feature importance method, the relative importance of the features in the logistic regression model is decomposed and the SHAP histogram is plotted, and the SHAP of the logistic regression model before and after the practice is shown in Fig. 5 and Fig. 6.

As can be seen from Figure 5 and Figure 6, among the top 4 indicators with the strongest explanatory strength in the Logistic regression model before strategy implementation, the student organization construction indicator, the system construction indicator, the party member development and education indicator and the propaganda demonstration indicator. This indicates that these 4 major categories of indicators in the Logistic regression model, all have an important contribution to the prediction of the effect of the implementation of the strategy, which in the top 4 indicators with the strongest explanatory strength, student organization construction indicators, indicating that in the pre-strategy implementation Logistic regression model, the student organization construction indicators as a whole have the most prominent explanatory strength for the output results of the model. Among the top 4 indicators with the strongest explanatory strength in the post-strategy implementation Logistic regression model, the student organization construction indicator, the system construction indicator, the party development and education indicator, and the propaganda demonstration indicator. As with the pre-epidemic logistic regression model, all four major categories of indicators have a significant impact on the model output in the post-strategy implementation logistic regression model. In addition to this, the 4 indicators of student organization building indicators, system building indicators, party member development and education indicators and propaganda demonstration indicators are among the top 4 indicators with the strongest explanations in both the pre-strategy and post-strategy logistic regression models.

4

Conclusion

The article explores and researches the digital transformation of party building in higher vocational colleges through the performance evaluation system of party building in higher vocational colleges, which is combined with a logistic regression model.

The article draws several conclusions through empirical research: 1)

The Logistic regression model constructed in this article has good validity and stability, and the prediction accuracy of the Logistic model on the overall sample group sample of a higher vocational college in 2023 is 84.6%.

2)

In terms of the explanatory strength of the indicators of the Logistic regression model, the top four indicators with the greatest contribution to the output results of the model are the indicators of student organization construction, system construction, party member development and education, and propaganda and demonstration, among which the overall student organization construction indicators are the most prominent in terms of their explanatory strength for the output results of the model.

Idioma:: Inglés

Calendario de la edición:: 1 veces al año
Temas de la revista:: Ciencias de la vida, Ciencias de la vida, otros, Matemáticas, Matemáticas aplicadas, Matemáticas generales, Física, Física, otros

RSS Feed de revista

Strategies and Practical Exploration of Digital Transformation of Party Building Work in Higher Education Institutions

Yuyao Wu

Publicado en línea: 19 mar 2025

Recibido: 09 nov 2024

Aceptado: 18 feb 2025

DOI: https://doi.org/10.2478/amns-2025-0468

Palabras claveLogistic regression model, Party building in higher vocational colleges, Assessment of indicator contribution, Digital transformation

© 2025 Yuyao Wu, published by Sciendo

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

Palabras clave
Logistic regression model, Party building in higher vocational colleges, Assessment of indicator contribution, Digital transformation