Research on the Path to Improving the Teaching Effectiveness of News Communication in Colleges and Universities under the Background of Artificial Intelligence

With the rapid development of artificial intelligence technology, all walks of life around the world are exploring the application potential of artificial intelligence (AI), and the field of education is no exception. AI not only changes the way we acquire knowledge but also promotes innovation in teaching concepts and methods. In university education, the application of AI technology provides more intelligent and personalized support for teaching. Especially in journalism and communication, AI technology has injected new vitality into the traditional teaching mode [1]. Through intelligent tutoring systems, data analysis platforms, virtual reality, and other technologies, educators can provide students with more accurate and efficient teaching services and, at the same time, bring new opportunities for classroom teaching and subject development.

The discipline of journalism has the characteristics of paying equal attention to theory and practice. Its teaching content not only includes teaching fundamental theories but also involves the skill cultivation of journalism practice, data processing, and content creation. The traditional education mode of journalism communication is based on classroom teaching, which depends on teachers' experience and students' active learning. With the rapid development of information technology, especially the acceleration of digitalization and intelligence, traditional teaching methods have gradually exposed some problems, such as the difficulty of teaching content to keep up with the development of the times and the inability of students' learning to be personalized [2]. Therefore, breaking through the limitations of traditional education and improving the teaching effect have become vital problems that need to be solved urgently in college journalism communication education.

The development of AI technology provides new possibilities for the reform of journalism communication education. AI technology can effectively improve the teaching effect through personalized recommendations of teaching content, intelligent evaluation feedback, and virtual laboratory construction [3]. For example, the intelligent learning system can automatically adjust the course content through data analysis and provide personalized push according to different students' learning progress and interests. The intelligent evaluation system can monitor students' learning status in real time and help teachers find students' learning bottlenecks in time to provide targeted counseling. In addition, virtual and augmented reality technologies also have excellent application prospects in the experimental teaching of news dissemination, which can provide students with a more intuitive and immersive news production and dissemination experience.

Artificial intelligence technology has shown great potential for application in education. However, the application of AI still needs specific challenges and limitations in journalism communication teaching in colleges and universities [4]. How to scientifically and effectively integrate AI technology into the teaching process and improve the teaching effect has become an urgent problem for educators and academic circles. This article explores how to use artificial intelligence to improve teaching effects by investigating and analyzing the current status of AI applications in journalism and communication majors in 30 universities across the country. The core purpose of the research is to propose feasible AI application paths, optimize teaching methods, and provide a reference for teaching reform in other disciplines.

2

Overview of related theories and technologies

2.1

Artificial intelligence empowers news dissemination

Artificial intelligence's strengthening of news timeliness is multi-subject. In the past, news dissemination mainly relied on mass media. Specifically, it relied on journalists and editors working in mass media organizations. Under the paradigm of journalism professionalism, a few elites undertake the social task of news dissemination. With the advent of the Internet era, especially the mobile internet era, user-produced content (UGC) and professional-produced content (OGC) are competing. In the era of artificial intelligence, robot production content (MGC) has also joined in [5]. Due to the diversification of production entities, it makes up for the time delay problem caused by journalists' energy limitations. The time delay is because journalists can only be present for some of the news, and they can only deal with multiple news clues one by one. In the situation of diversified production entities, journalists have more sources of information, the audience's identity is excessive compared to that of producers and sellers, and the status of machine production content is becoming more and more prominent. The ternary production subject accelerates the efficiency of news dissemination.

Artificial intelligence produces news without interruption [6]. For example, sensors can collect information 24 hours a day, writing robots can write news endlessly, and AI anchors can broadcast news without interruption.

The impact of artificial intelligence on news authenticity still needs more discussion. On the one hand, due to the diversification of production entities and the extensive channels for people to get news, users can directly get in touch with more first-hand information instead of seeing a more real world in the mimic environment created by the media through the selection of information and the setting of the agenda [7]. However, fake news has become the most prominent problem in this era. The use rate of fake news has increased by 365% since 2016, and it became the hot word of the year in the Corinthian Dictionary in 2017. "The reason is the loss of journalistic professionalism and the lack of users' media literacy. Most producers of UGC content have yet to receive professional news training, cannot distinguish fake information sources, and, what is more, they will use fake news to create gimmicks to attract traffic. Fake news, coupled with the algorithm distribution of artificial intelligence, has amplified the spread power of fake news [8]. Users need more media literacy, forwarding a large number of fake news, and the distribution system of the algorithm makes the algorithm misjudge this fake news as important news, which increases the distribution of these news.

Regarding the proximity and interest of news, the algorithm also positively affects these two news value elements because the algorithm distribution mechanism lies in the collection of user behavior data to judge users' preferences and push news that they are interested in to users on this basis.

2.2

An overview of journalism communication teaching in colleges and universities under the background of artificial intelligence

With the rapid development of artificial intelligence technology, more and more colleges and universities have begun to explore the application of AI technology to the teaching reform of journalism and communication [9]. The introduction of AI technology provides brand-new teaching methods and methods for journalism communication education in colleges and universities, especially showing great potential in personalized teaching, intelligent evaluation, and interactive learning. AI can help teachers accurately grasp students' learning needs and progress through extensive data analysis and intelligent algorithms, as well as achieve personalized teaching and realtime feedback, thereby improving the effectiveness and pertinence of teaching. At the same time, AI can also improve students' learning experience, enhance classroom interaction and participation, and promote cultivating students' comprehensive ability and innovative thinking.

The traditional teaching mode of news communication mainly relies on classroom teaching, case analysis, and practical courses. It pays attention to the imparting of theoretical knowledge and the cultivation of practical operation ability. However, with the rapid development of digitalization and information technology, traditional teaching has gradually exposed some limitations, especially in better meeting students diversified and personalized learning needs [10]. The introduction of AI can provide accurate data analysis and evaluation tools and help students choose the most suitable learning path according to their personal interests and learning progress through adaptive learning platforms and intelligent auxiliary tools, thereby improving the depth and effect of learning.

In the teaching of journalism and communication majors, the application of AI is not only limited to teaching aids but also plays an active role in curriculum design, practical teaching, data analysis, etc [11]. For example, the intelligent learning management system can help teachers track students' learning progress and knowledge mastery in real-time, automatically identify the difficulties encountered by students in the learning process, and provide targeted tutoring programs. With the introduction of virtual and augmented reality technology, the experimental teaching of journalism communication has been greatly enriched. Students can simulate the whole news production process through virtual scenes, experience the operation of different news production positions, and then improve their practical operation ability and news sensitivity. Although artificial intelligence has brought many positive changes and innovations to journalism communication education in colleges and universities, it also faces some challenges and problems in its practical application.

3

Construction of Teaching Framework of Journalism Communication in Colleges and Universities Based on Artificial

3.1

Model framework

Aiming at the problem of improving journalism communication teaching in colleges and universities this paper puts forward a multi-sector growth model, which includes teaching departments and demand departments, aiming to solve the problem of improving journalism communication teaching in colleges and universities through artificial intelligence [12]. The multi-sectoral growth model has become an essential tool for studying the transformation of educational structure.

The modeling of the multi-sector growth model mainly constructs a teaching effect evaluation model based on artificial intelligence assistance, aiming at analyzing how education departments can improve the teaching quality of journalism and communication disciplines through intelligent means [13]. The model assumes that the education sector includes three modes: traditional teaching, intelligent teaching, and hybrid teaching, and the impact of each mode on teaching effectiveness can be optimized through data-driven analysis.

The key to modeling teaching departments is to characterize the new generation of artificial intelligence in teaching. There are two main methods in the existing literature to describe the application of artificial intelligence in education: one is the teaching model based on task automation, and the second is the technology-enhanced teaching reform model. The teaching model based on task automation regards artificial intelligence as the latest stage of teaching process automation [14]. Artificial intelligence is replacing some teachers' functions through technical means. Based on the task automation model, the degree of application of artificial intelligence in teaching is measured by quantifying the proportion of intelligent teaching tasks in all tasks. The technology-enhanced teaching reform model regards artificial intelligence as enhancing and optimizing teaching technology. The introduction of artificial intelligence technology improves the quality and efficiency of education, and then the development of teaching reform is promoted [15]. This paper regards the new generation of artificial intelligence as a technology-enhanced teaching reform and improves the teaching effect through intelligent teaching tools and platforms. It describes the application of the new generation of artificial intelligence in education through intelligent teaching resources and platforms. In order to highlight the difference between intelligent teaching tools and traditional educational means, this paper introduces both traditional and intelligent teaching methods.

First, the education system's teaching departments can be divided into two categories: theoretical teaching and practical teaching, distinguished by the subscript j = (t, p), where j = t represents theoretical teaching and j = p represents practical teaching [16]. There are many identical educational institutions in the education system, and representative educational institutions make each sector a perfectly competitive market for teaching decisions. Representative educational institutions of theoretical and practical teaching all adopt constant substitution flexibility technology for teaching design, and the specific function form is shown in (1).

1

E = α_{1} \cdot A + α_{2} \cdot C + α_{3} \cdot S

Among them, E represents the teaching effect, A represents students' academic performance, C represents students' classroom participation, S represents the attractiveness of teaching content, and α₁, α₂, and α₃ represent the weight coefficients of each factor [17]. This shows that both intelligent infrastructure and traditional infrastructure will affect the production functions of manufacturing and service sectors, and the impact on the production technology of these two sectors is unbiased.

If P_{jt, rt}, and w_t represent the price of courses provided by the teaching department in the t-th period, the capital rent of teaching resources and the salary of teachers, respectively, and solve the problem of maximizing the teaching effect of representative educational institutions, the first-order optimal condition of the equilibrium of the education department can be obtained [18]. The specific solution process is as follows. First, the constrained extremum equation is constructed, as shown in Equation (2).

2

\begin{matrix} \max H = P_{j t} Q_{j t} - (r_{t} K_{j t} + w_{t} L_{j t}) \\ s . t . Q_{j t} = A_{j} G_{t}^{γ} {[α_{j}^{1 / σ_{j}} {(B_{j} Z_{t}^{ϕ} K_{j t})}^{(σ_{j} - 1) / σ_{j}} + {(1 - α_{j})}^{1 / σ_{j}} {(L_{j t})}^{(σ_{j} - 1) / σ_{j}}]}^{\frac{σ_{j}}{σ_{j} - 1}} \end{matrix}

Then, according to the constrained extremum equations, the partial derivatives of K_jt and L_jt are solved, respectively. The respective partial derivatives are equal to zero, and the relative demand functions of K_jt and L_jt satisfying the condition of maximizing the producer's profit can be obtained. Specifically, it is shown in formulas (3) and (4).

3

P_{j t} \cdot {(A_{j} G_{t}^{γ})}^{(σ_{j} - 1) / σ_{j}} \cdot {(Q_{j t})}^{(σ_{j}} \cdot α_{j}^{1 / σ_{j}} \cdot {(B_{j} Z_{t}^{ϕ} K_{j t})}^{(σ_{j} - 1) / σ_{j}} = r_{t} K_{j t}

4

P_{j t} \cdot {(A_{j} G_{t}^{γ})}^{(σ_{j} - 1) / σ_{j}} \cdot {(Q_{j t})}^{1 / σ_{j}} \cdot {(1 - α_{j})}^{1 / σ_{j}} \cdot {(L_{j t})}^{(σ_{j} - 1) / σ_{j}} = w_{t} L_{j t}

The model divides the teaching departments into three production departments: private education investment, traditional teaching infrastructure investment, and intelligent teaching infrastructure investment, distinguished by subscript n = (k, g, z), respectively. k represents the private education investment sector, g represents the traditional teaching infrastructure production sector, and z represents the intelligent teaching infrastructure production sector [19]. The above three teaching investment departments are all input by representative educational institutions with theoretical and practical teaching output. In a perfectly competitive market, they constantly substitute flexible technology to produce educational services. The production function is shown in Equation (5).

5

I_{n t} = {[ω_{n}^{1 / ε_{n}} {(I_{n m t})}^{(ε_{n} - 1) / ε_{n}} + {(1 - ω_{n})}^{1 / ε_{n}} {(I_{n s t})}^{(ε_{n} - 1) / ε_{n}}]}^{\frac{ε_{n}}{ε_{n} - 1}}

The subscript t represents time, and I_nt represents the educational service output. I_nmt and I_nst of the education investment department in the t-th period n, respectively, represent the input amount of theoretical teaching and practical teaching services in the education investment production department in the t-th period n [20]. The parameter ε_n is a constant, and its value range is (0, ∞), indicating the substitution elasticity of theoretical and practical teaching services in n education investment departments. The parameter w_n is a constant with 0 < w < 1 value range.

Similarly, if P_nt, P_mt, and P_st represent the market prices of the t-th education investment service, theoretical teaching content, and practical teaching content, respectively, where n = (k, g, z) represent different education investment departments [21]. By solving the problem of maximizing the teaching effect of representative educational institutions in the production department of educational investment products, the first-order optimal conditions for maximizing the profits of educational institutions in teaching investment can be obtained. The specific solution process is shown in formula (6).

6

\begin{matrix} \max J = P_{n t} I_{n t} - (P_{m t} I_{n m t} + P_{s t} I_{n s t}) \\ s . t . I_{n t} = {[ω_{n}^{1 / ε_{n}} {(I_{n m t})}^{(ε_{n} - 1) / ε_{n}} + {(1 - ω_{n})}^{1 / ε_{n}} {(I_{n s t})}^{(ε_{n} - 1) / ε_{n}}]}^{\frac{ε_{n}}{ε_{n} - 1}} \end{matrix}

The demand department is an indispensable part of news communication teaching, and each student has the same learning preference parameters; that is, they face the same learning resource acquisition conditions, the same learning background, and the same learning development opportunities. Students earn knowledge income by participating in classroom learning, providing academic achievements to obtain academic achievement rewards, purchasing educational resources for learning, and accumulating learning experiences for future development [22]. The demand department in the model is characterized by a representative group of students who exist indefinitely. Students make intertemporal choices in a perfectly competitive learning environment, pursuing the maximization of their lifelong academic utility. Generally speaking, the form of the academic utility function of representative students is shown in (7).

7

U = \sum_{t = 0}^{\infty} β^{t} \frac{C_{t}^{1 - ρ}}{1 - ρ}

Among them, β and ρ are constants, and the value ranges are 0 <β< 1 and ρ > 0, respectively. C denotes students' learning outcomes in period t, and β denotes the time preference factor. ρ is the critical parameter of the student utility function, determining students' willingness to adjust their learning input in different periods [2]. When the value of β is smaller, the decline of marginal utility is caused by increasing the current learning input. Hence, students are more inclined to choose the change of intertemporal learning input. It is not difficult to prove that when β → 0, the utility function is approximately a linear function of C, and students will obtain higher learning benefits by considerably adjusting the intertemporal learning investment. Therefore, ρ can also be regarded as the reciprocal of the substitution elasticity of students' intertemporal learning investment; β has nothing to do with C, that is, the substitution elasticity between any two learning inputs is equal, and C is the compound learning outcome, which satisfies the following conditional formula as shown in (8).

8

C_{t} = {[ω_{c}^{1 / ε_{c}} {(C_{m t})}^{(ε_{c} - 1) / ε_{c}} + {(1 - ω_{c})}^{1 / ε_{c}} {(C_{s t})}^{(ε_{c} - 1) / ε_{c}}]}^{\frac{ε_{c}}{ε_{c} - 1}}

Where in the subscript t represents time, C_t represents the compound learning result of the t-th period; C_mt and C_st represent the quantity of theoretical and practical teaching content used for learning in the t-th period, respectively [23]. The parameter ε is a constant, and its value range is (0, ∞), indicating the substitution elasticity of theoretical and practical teaching content in the education market. The parameter w_c is a constant whose value range is 0 < w_c < 1.

Students are competitive, and each student faces a given unit academic reward rate and learning resource interest rate. The model sets the initial learning capital K0, which is strictly greater than zero. In each period, the amount of learning capital students hold is Kt, and they will get the income of learning capital risk [24]. At the same time, the number of labors provided is Lt, and the learning reward income obtained is r_tK_t. Therefore, the total income obtained by a representative student is the sum of learning capital gains and learning reward income, which is rk + wl. This paper introduces the education management department into the model, and the model simplifies the education management department. It is assumed that the total amount of education tax collected by the government in each period is T, and all the tax revenue is used for traditional teaching infrastructure and intelligent teaching infrastructure expenditure, so there is no fiscal deficit. Students use part of their total after-tax income for learning consumption and the rest for academic savings. The savings from private education investment and the investment is used to accumulate more learning capital. The total amount of private education investment in the current period minus capital depreciation is equal to the increment of learning capital in the current period. It is a state variable in the model. From this, the student's learning budget constraint equation can be obtained, as shown in formula (9): 9 ${\begin{matrix} P_{m t} C_{m t} + P_{s t} C_{s t} + P_{k t} I_{k t} + T_{t} = r_{t} K_{t} + w_{t} L_{t} \\ K_{t + 1} = (1 - δ^{k}) \cdot K_{t} + I_{k t} \\ K_{0} > 0 \end{matrix}$

The parameter δ^k is a constant, representing the depreciation rate of learning capital, and its value range is δ^k ∈ (0, 1). In order to further simplify the solution of the model, this paper introduces the finite period T. It transforms the utility maximization model under the indefinite existing representative student budget constraint into the utility maximization model under the finite period T student budget constraint. Then, solve the optimal condition of maximizing student utility with limited time, and analyze the time limit in the optimal condition. By finding the limit of T, we can get the optimal condition of maximizing the utility of students who exist indefinitely.

3.2

Mathematical model analysis

The goal of mathematical model analysis is to clarify the influence mechanism of the new generation of artificial intelligence on the teaching effect of journalism communication in colleges and universities, that is, to explore the marginal effect of increasing intelligent teaching infrastructure on optimizing teaching effect. Generally speaking, theoretical models' comparative static equilibrium analysis results are also applicable to dynamic equilibrium models. In order to make the expression more intuitive, the time subscript t of variables omitted in static analysis is compared.

The analysis of the critical path to improve the teaching effect of journalism communication in colleges and universities is shown in Figure 1. As can be seen from the figure, the path to improving the teaching effect of news communication in colleges and universities can be carried out through two levels: the optimization of internal teaching structure and the optimization of inter-teaching structure. First of all, by analyzing the change in input intensity of teaching elements, the distribution ratio of knowledge and technology should be reasonably adjusted; for example, the proportion of technology in intelligent teaching should be increased to 70%, and the proportion of knowledge in traditional teaching should be maintained above 60%, to optimize the internal structure of teaching. Secondly, it promotes the gradual increase of intelligent teaching overall. For example, the participation rate of intelligent teaching will be increased from 35% to 60% within five years, and the inter-teaching structure will be dynamically optimized. Combining intelligent data analysis and realtime feedback mechanisms, the optimization path and strategy can be further clarified, and a scientific basis can be provided to improve the teaching effect.

There are two teaching departments in the education system, traditional teaching and intelligent teaching, which use two teaching elements: knowledge and technology. Teaching effect optimization includes two connotations: The internal structure optimization of teaching and the structural optimization of teaching. The optimization of the internal structure of teaching is mainly manifested in the process of reconfiguring teaching elements. In this paper, the change in element density of teaching departments is used to reflect the level of optimization of the teaching effect, and the change in the ratio of knowledge and technology input among teaching departments is used to reflect the direction of element flow. Structural optimization between teaching and the relative structural change between traditional teaching and intelligent teaching usually refers to the relative proportion of traditional teaching and intelligent teaching or the change of the proportion of intelligent teaching.

Therefore, x^k and x^l are defined in this paper, respectively, as the proportion of knowledge input and technology input used by the traditional teaching m department and are used to measure the intensity of knowledge and technology used by the teaching department, as shown in the formula (10). Without losing generality, it is assumed that x^k> x^l. The traditional teaching department is knowledge-intensive, while the intelligent teaching department is technology-intensive.

10

{\begin{matrix} x^{k} = K_{m} / K \\ x^{l} = L_{m} / L \end{matrix}

From this, K_m = x^kK, L_m = xlL, K_s = (1-x_k), K, and L_s = (1-x_l) L. This section needs to clarify two questions: 1)

Will the new generation of artificial intelligence change the element intensity of traditional or intelligent teaching?

2)

Whether the new generation of artificial intelligence will change the relative proportion of traditional teaching and intelligent teaching?

To address these questions, we conducted a series of studies. Studies have shown that, on the one hand, the optimization of the teaching effect is influenced by supply-side factors, such as technological progress and the increase of educational resources, which will lead to the evolution of the teaching structure. On the other hand, it is influenced by demand-side factors, such as the demand changes of different teaching modes and teaching contents. Then, it affects the optimization and upgrading of the teaching effect. Therefore, this paper analyzes the mechanism of the new generation of artificial intelligence's influence on optimizing the teaching effect from two levels of supply and demand. When discussing the influence mechanism of the supply side, simplify the model setting of the demand side (Hypothesis 1), that is, control the influence of changes in demand factors on the teaching effect and only discuss the marginal effect of the supply side. When discussing the mechanism of demand-side influence, the model setting of supply-side is simplified (Hypothesis 2), that is, the influence of changes in supply factors on teaching effect is controlled, and only the marginal effect of demand-side is discussed. The specific assumptions are as follows:

Hypothesis 1: The structural proportion of representative students choosing traditional teaching and intelligent teaching resources is the same as the structural proportion of traditional teaching and intelligent teaching resources invested by educational investment and production departments, which are equal to constants w and 1-w. In the students and educational investment production sectors, the substitution elasticity of traditional teaching and intelligent teaching resources is the same, both equal to the constant ε.

Hypothesis 2: In the production functions of traditional teaching and intelligent teaching resources, the elasticity of knowledge and technology substitution is equal to constant 1. The parameters αj in the production functions of traditional teaching and intelligent teaching resources are equal to the constant α.

When Hypothesis 1 is established, it means that the proportion of representative students and the three educational investment and production departments using or investing in traditional teaching and intelligent teaching resources is the same, and the relative proportion of market demand for traditional teaching and intelligent teaching resources remains unchanged. Therefore, the demand structure of traditional teaching and intelligent teaching resources will not affect the optimization of teaching effect. Therefore, according to the first-order condition of maximizing the utility of the representative education sector, formula (11) can be obtained.

11

\frac{P_{m} Q_{m}}{P_{s} Q_{s}} = \frac{ω}{1 - ω} \cdot {(\frac{P_{m}}{P_{s}})}^{1 - ε}

Among them, P_m represents the quality of information feedback, Q_m represents technical accuracy, P_s represents response speed, Q_s represents participation, ω represents weight coefficient, and ε represents classroom participation.

As the number of intelligent infrastructures continues to increase, the development level of the new generation of artificial intelligence technology continues to improve, and educational resources and teaching methods will be reconfigured between traditional teaching and intelligent teaching. The influence direction of the new generation of artificial intelligence on the factor intensity of teaching departments depends on the elasticity difference of knowledge and technology output between traditional teaching and intelligent teaching, as well as the substitution elasticity of traditional teaching and intelligent teaching resources and the substitution elasticity difference of knowledge and technology in production departments. The influence of the new generation of artificial intelligence on the relative proportion of traditional teaching and intelligent teaching depends on the difference in knowledge output elasticity between traditional teaching and intelligent teaching and the substitution elasticity of traditional teaching and intelligent teaching resources. The influence of the new generation of artificial intelligence on the relative proportion of traditional teaching and intelligent teaching output (the actual output of traditional teaching) depends on the difference in knowledge output elasticity between traditional teaching and intelligent teaching. The direction of the new generation of artificial intelligence influences the ratio of knowledge and technology income in traditional teaching and intelligent teaching, depending on the substitution elasticity of knowledge and technology within traditional teaching and intelligent teaching. The following two exceptional cases discuss the influence mechanism behind this phenomenon specifically.

The integration process of intelligent teaching tools and traditional teaching methods is shown in Figure 2, from which it can be seen that the integration of intelligent teaching tools and traditional teaching methods is one of the critical paths to improve the teaching effect of news communication in colleges and universities. First of all, through the survey data, it is found that at present, 75% of journalism communication courses still need to be dominated by traditional teaching methods, and only 25% of courses have introduced intelligent teaching tools. In order to improve the integration effect, the hybrid teaching mode of "intelligent assistance + traditional teaching" can be adopted. For example, intelligent tools can be used in news writing courses for real-time grammar detection and data analysis. In contrast, traditional lectures can provide creative thinking and practical guidance. Experimental data show that in the classroom where intelligent tools are introduced, students' average homework completion time is shortened by 30%, and the error rate is reduced by 25%. It is planned to increase the application proportion of intelligent tools in teaching to more than 50%. Phased training and teacher adaptation plans will ensure the deep integration of traditional methods and intelligent tools and provide students with a more efficient and personalized learning experience.

4

Experiment and Results Analysis

The test of autoregressive (AR) model is a commonly used model detection method in teaching. AR (1) is used to check whether the first-order autoregressive term in the model is significant, and AR (2) test is used to check whether the second-order autoregressive term in the model is significant. This paper constructs an econometric model to analyze the related experiments of artificial intelligence to improve the teaching effect of news communication in colleges and universities. In order to improve the effectiveness of estimation, the systematic moment estimation method is used to solve the endogenous problem. Experiments in this paper show that artificial intelligence and teaching effects are causally related, which may lead to endogenous problems. In order to avoid the error of least square estimation, this paper uses the lag term as the instrumental variable and chooses the robust standard error. The system moment estimation method needs an autocorrelation test and an over-identification test. If the AR (1) test result is significant, there is an endogenous problem; If the AR (2) test result is insignificant, the model has no endogenous problem. The Hansen test is used to determine whether the instrumental variable is valid. If the null hypothesis is not rejected at the significance level of 10%, the instrumental variable is valid. This paper uses this method to analyze national and regional sample data empirically. The results of the autocorrelation and overidentification tests of the system moment estimation method are shown in Table 1.

Table 1.

Results of Autocorrelation Test and Over-identification Test of System Moment Estimation Method

Teaching mode	Traditional teaching mode	Intelligent teaching mode	AR (1)	AR (2)
Student satisfaction	72%	88%	0.020	0.267
Improvement of learning efficiency	65%	82%	0.032	0.152
Teachers' participation in teaching	60%	78%	0.017	0.187
Publication of research results	30%	22%	0.032	0.214

It can be seen from the table that the AR (1) test result is 0.032 (p-value < 0.05), indicating that there is a first-order serial correlation, so the null hypothesis is rejected, indicating that the model may have endogenous problems. However, the result of the AR (2) test was 0.187 (p-value > 0.05), indicating that there was no second-order sequence autocorrelation, so no second-order autocorrelation problem was found, and the model endogeneity was solved. Further, the Hansen test result is 0.214 (p-value> 0.10), which cannot reject the null hypothesis of the validity of instrumental variables, indicating that the selected instrumental variables are valid. These test results provide a reliable basis for subsequent empirical analysis.

The graph of students' average grade data is shown in Figure 3. It can be seen that in the experimental group, the average score of students who use artificial intelligence-assisted teaching improved significantly. Specifically, the average score of students in the traditional teaching group was 68.4 points. In comparison, the average score of students in the experimental group was 81.7 points, an increase of 13.3 points, or about 19.4%. Especially for students with lower grades, the improvement in performance after using artificial intelligence tools is more prominent. The average score of students with low grades has increased by 16.8 points, showing the potential of artificial intelligence in promoting students' personalized learning and improving academic performance. At the same time, students' satisfaction with intelligent teaching is high, with a score of 4.5/5, further proving the positive effect of artificial intelligence tools on improving teaching effects.

Figure 4 has showed the classroom Engagement Data. The classroom engagement data graph is shown in 4. According to the data graph, classroom participation is significantly different between the experimental group (class using artificial intelligence-assisted instruction) and the control group (class with traditional instruction). Specific data showed that the classroom participation rate of students in the experimental group was 85.2%, while that in the control group was 65.4%, which was 19.8 percentage points higher than that in the control group. In addition, the number of students' questions and the frequency of classroom interactions also increased. The average number of questions per class in the experimental group was 12, compared with 6 in the control group, and the frequency of interactions increased by 100%. Especially in group discussions and online interactive sessions, students' participation and enthusiasm for discussion are higher after using artificial intelligence tools, showing the positive effect of intelligent teaching on stimulating students' classroom participation.

It can be seen from the Table 2 that the use of artificial intelligence tools at different frequencies has improved the teaching effect of journalism communication courses. The higher the frequency of use, the significantly improved the teaching effect in terms of learning initiative, knowledge mastery, interaction, and timeliness of feedback.

Table 2.

The Impact of Artificial Intelligence Tools on The Teaching Effect of Journalism Communication Courses

Frequency of use of AI tools	1 hour per week	3 hours per week	5 hours per week	Effect Increase (%)
Student learning initiative	65%	76%	85%	30%
Mastery of curriculum knowledge	72%	81%	90%	25%
Course interactivity	60%	70%	80%	33%
Timeliness of after-school feedback	58%	75%	85%	47%

The comparison chart of teaching satisfaction is shown in Figure 5. It can be seen from the figure that the satisfaction of students in the experimental group with the course content is 90.5%, which is significantly higher than that of the control group (75.3%), an increase of 15.2 percentage points. Regarding teaching interactivity, the satisfaction of the experimental group was 88.7%, while that of the control group was 70.6%, which was 18.1 percentage points higher. These data show that artificial intelligence-assisted instruction has significantly improved students' satisfaction with course content, teaching interaction, and overall teaching effect, especially regarding course interaction and content innovation.

The comparison chart between AI and educational content data is shown in Figure 6. It can be seen from the figure that in terms of the update frequency of course content, the experimental group (AI-assisted teaching) updates the content 12 times per semester. In comparison, the control group (traditional teaching) updates the content 4 times. The update frequency of the experimental group is 3 times that of the control group. As for the survey of students' interest in course content, the interest score of the experimental group was 85.3%, which was significantly higher than that of the control group's 68.7% and increased by 16.6 percentage points. Overall, the overall satisfaction score of the experimental group in educational content was 88.9%. In comparison, that of the control group was 72.3%, and the overall satisfaction of the experimental group was 16.6 percentage points higher.

The relationship between teaching effect and student achievement is shown in Figure 7. It can be seen from the data chart that from the perspective of teaching effect scores, the teaching effect score of AI-assisted teaching classes is 88.7%, while that of traditional teaching classes is 72.5%. This gap reflects the significant improvement of artificial intelligence in teaching effectiveness. The proportion of students in the high segment (above 90 points) in AI classes is 25.7%, while that in traditional classes is only 14.3%. Regarding the proportion of students in low segments (below 60 points), the AI class is only 4.2%, and the traditional class is 12.5%. This shows that artificial intelligence-assisted teaching improves the overall performance level and effectively reduces the proportion of low-performance students.

The relationship between students' learning outcomes and feedback timeliness is shown in Figure 8. The data show a strong correlation between the timeliness of feedback and students' learning achievements, and timely feedback can significantly improve students' academic performance and satisfaction, especially in low-segmented students. The improvement of feedback timeliness significantly reduces the proportion of low-grade students. Therefore, the fast feedback mechanism assisted by artificial intelligence is a critical factor in improving students' learning outcomes. The influence of feedback timeliness on students with low segments is particularly prominent: among students who give timely feedback, the proportion of students with scores below 60 is 4.1%, while in the group with slower feedback, this proportion is 12.3%.

Figure 9 has showed the relationship between students' self-study time and achievement changes. The data shows that the average score of students who spend less than 1 hour of self-study every day is 72.5 points; The average score of students with self-study time of 2 hours is 80.3 points, an increase of 7.8 points, an increase of about 10.8%. Students' self-study time significantly affects their academic achievements, especially students whose self-study time is more than 2 hours, and their achievements are improved. Especially among low-grade students, increasing self-study time can significantly reduce the proportion of low-grade students. Therefore, encouraging students to increase their selfstudy time is an effective way to improve learning effects. Using artificial intelligence auxiliary tools also helps students extend their self-study time and improve their learning results.

The analysis of students' achievement changes under different teaching modes is shown in Table 3. It can be seen from the table that compared with the artificial intelligence-assisted teaching mode, the average score of students has increased to 83.5 points, an increase of 9.3 points, an increase of about 12.5%. In the artificial intelligence-assisted mode, the distribution of student achievements is more concentrated, and the proportion of students with scores above 80 has increased from 42% in the traditional mode to 72%. Data analysis shows that intelligence-assisted teaching has significantly improved students' overall grades and learning performance. Especially in the proportion of students in high segments, the performance of artificial intelligence teaching mode is more prominent, and it can effectively reduce the proportion of students in low segments. Based on these results, it can be speculated that artificial intelligence teaching mode can better stimulate students' learning enthusiasm and improve their academic level. It is a crucial way to improve the teaching effect of journalism and communication courses.

Table 3.

Analysis of Student Achievement Changes under Different Teaching Modes

Teaching mode	Average grade	Achievement improvement
Theoretical knowledge mastery	75	13.33%
Improvement of practical ability	70	14.29%
Information acquisition efficiency	68	14.71%
Demonstration of innovation ability	72	13.89%

5

Conclusion

Under the background that artificial intelligence technology increasingly penetrates the field of education, the improvement of the effect of journalism communication teaching in colleges and universities has shown significant changes. Through the analysis of the survey data of journalism and communication majors in 30 universities across the country, this paper discusses the application and practical path of AI technology in the teaching of this subject and draws the following conclusions: 1)

According to survey data, students' academic performance in colleges and universities that use AI technology to assist teaching has generally improved. Specific data show that the average grade of these college students has increased by about 15%. Among them, using personalized learning systems and intelligent teaching evaluation tools can accurately grasp students' learning progress and difficulties to adjust teaching content, help students overcome learning difficulties, and improve learning effects. For example, under the traditional teaching mode, the average score of students in journalism and communication courses in a university is 75 points. However, after the introduction of AI assistance, the average score reaches 86 points. In addition, classroom participation has also increased significantly, from 62% in the traditional mode to 83%. This change shows that AI technology can significantly increase students' learning initiative and participation by enhancing classroom interactivity.

2)

AI provides students a personalized learning experience and dramatically optimizes teachers' teaching process. The survey shows that after teachers use AI for teaching management and data analysis, teaching efficiency has increased by more than 20%. Teachers can monitor student's learning progress and problems in real-time through the AI-assisted learning management system and adjust teaching content and methods in time. Further data analysis shows that introducing AI has reduced teachers' lesson preparation time by an average of 30% while increasing interaction time in class by 25%. This improvement in efficiency enables teachers to focus more on the deepening of teaching content and the personalized guidance of students.

3)

The experimental teaching of journalism and communication majors is efficient and operational, and the traditional teaching mode is often challenging and requires more simulation and practice opportunities. Students can practice news production and dissemination in simulated environments through AI technologies, primarily virtual and augmented reality. This immersive experience improves students' practical skills and the intuition and fun of teaching. According to the survey data, students' satisfaction with experimental teaching in colleges and universities using VR and AR technology has increased by about 25%, and the qualification rate of participating in students' professional skills tests has increased by 18%. For example, a university's experimental pass rate in traditional news reporting courses is 70%, but after the introduction of VR technology, the pass rate increases to 88%, which shows that AI technology can effectively improve students' hands-on ability and practical ability.

Język:: Angielski

Częstotliwość wydawania:: 1 razy w roku
Dziedziny czasopisma:: Nauki biologiczne, Nauki biologiczne, inne, Matematyka, Matematyka stosowana, Matematyka ogólna, Fizyka, Fizyka, inne

Kanał RSS czasopisma

Research on the Path to Improving the Teaching Effectiveness of News Communication in Colleges and Universities under the Background of Artificial Intelligence

Jing Song

Data publikacji: 27 lut 2025

Otrzymano: 09 paź 2024

Przyjęty: 31 sty 2025

DOI: https://doi.org/10.2478/amns-2025-0118

Słowa kluczoweartificial intelligence, news communication, teaching effect, personalized learning

© 2025 Jing Song, published by Sciendo

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

Słowa kluczowe
artificial intelligence, news communication, teaching effect, personalized learning