Research on the Innovation of Artificial Intelligence Oriented Teaching Mode and Improvement of Experimental Teaching for Computer Science Majors
Published Online: Mar 17, 2025
Received: Nov 13, 2024
Accepted: Feb 13, 2025
DOI: https://doi.org/10.2478/amns-2025-0350
Keywords
© 2025 Haiyan Liu et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.
With the rapid development of science and technology, the level of artificial intelligence technology in China has developed rapidly, laying a good foundation for the realization of education and teaching informatization. The birth of artificial intelligence not only brings innovation to the traditional information technology industry but also changes the training mode of computer professional talents [1-3]. Nowadays, the training of computer professional talents no longer only focuses on programming ability but also needs to master AI thinking and innovation ability. The application of artificial intelligence will dominate future development, so computer professional education needs to transform from technical learning and training to innovative thinking, which is the direction led by artificial intelligence technology [4-6]. With the help of AI, computer majors can realize personalized education, provide precise teaching assistance for computer teachers, and help cultivate students’ AI thinking and innovation abilities [7-9]. Under the status quo of shortage of educational resources, students can be provided with richer educational resources through online educational platforms to help them better learn and master the application of computer technology and AI technology, and the rise and development of AI technology has had a profound impact on the education of computer majors and talent cultivation [10-12]. Therefore, it is necessary to create an AI-oriented teaching model for computer majors in order to achieve the innovation and optimization of the teaching concept as well as the teaching mode, to further stimulate the students’ learning interest, to enhance the efficiency of classroom teaching, and to promote the improvement of the comprehensive quality level of students [13-15].
This paper mainly focuses on the innovation of teaching mode and experimental teaching reform of computer specialties to carry out in-depth research. Through the improvement and optimization of the curriculum system, practical links, faculty, and intelligent evaluation mechanism, it promotes the smooth implementation of intelligent teaching mode innovation in computer majors. In the process of reforming the experimental teaching of computer majors, the experimental teaching system is constructed through artificial intelligence. The innovative teaching mode and experimental teaching were applied to practice, and the experimental class and control class were selected to verify the actual effect of the teaching mode on the cultivation of students’ computer literacy. Through a questionnaire survey, the students’ satisfaction with the intelligent experimental teaching is discussed.
With the rapid development of AI technology and the deepening of education informatization, smart education has become a new direction of educational change. Literature [16] summarizes the transformation of AI education experience into three paradigms: learner-receiver under the guidance of AI, learner-collaborator under the support of AI, and learner-leader under the empowerment of AI, and points out that the development trend of AI education is to enhance the autonomy and personalization of learners. Literature [17] found through experiments that most students recognize and like the teaching mode of computer application technology based on artificial intelligence, and the mode can improve students’ interest in learning computers, which plays an important role in the reform of computer teaching. Literature [18] puts forward the teaching concept of artificial intelligence + software to promote the intelligence and development of the computer profession and then cultivate applied and compound skilled talents for the whole society and local economic development, industrial transformation, and upgrading. Literature [19], in order to adapt the traditional computer teaching system to the rapidly developing artificial intelligence era, tries to apply artificial intelligence technology to the computer teaching system and then cultivates advanced applied talents useful to society and promotes social and economic development. Literature [20] took 45 undergraduate students who participated in a university-level programming course as research subjects to investigate the influence mechanism between programming education based on generative AI and students’ computational thinking ability, programming self-efficacy, and learning motivation, and the results showed that it is feasible to adopt generative AI technology in programming education, which can significantly improve students’ computational thinking ability, programming self-efficacy and classroom motivation. Literature [21] points out the key problems existing in the cultivation of computer professional talents at the present stage, discusses the construction of computer specialties under the background of artificial intelligence, and proposes a diversified talent cultivation model for computer specialties, aiming to better cultivate innovative talents in computer specialties adapted to the development of the times and oriented to the needs of applications.
The innovation of teaching mode in the context of artificial intelligence is shown in Figure 1, and the combination of theory and practice should be especially emphasized when designing related courses. From the requirements of professional certification of engineering education, the curriculum system should realize the support of graduation requirements. Taking the graduation requirements as an example, the supporting courses that should be set up are Introduction to Computer, High-level Language Programming, Object-Oriented Programming, Principles of Computer Composition, Algorithms and Data Structures, Computer Networks, Computer Operating Systems, Principles of Database Systems, Social Practice, and Innovative Practice.
Innovation of teaching mode in artificial intelligence
Colleges and universities should build a dual-creation teaching model for computer professional talent training and establish a multi-dimensional, innovative, practical training system consisting of different levels, such as in-class experiments, curriculum design, extracurricular practice, enterprise internships, and graduation design. Since many application demands in the field of computers come from society and enterprises, the in-depth integration between schools and enterprises in the field of AI and the Internet should be strengthened. A multi-channel, three-dimensional practical training system should be constructed with schools, online platforms, enterprise internship bases, etc.. The dual-creation practical links of joint internships inside and outside the school and synergistic integration of schools and enterprises should be constructed. Under the joint efforts of the Department of Computer Science and the employer, its talent cultivation aspect, the corresponding program will be formulated. However, in order to improve the student’s practical ability to really get exercise and constantly improve the student’s ability in this area, the school also needs to actively communicate with enterprises, and enterprises to jointly run schools and combine with the actual needs of enterprises, in the teaching content, to make the appropriate choice, and establish a good cooperative relationship with enterprises.
The cultivation of dual-creation talents of computer majors in the context of AI requires that teachers of computer majors introduce new knowledge and new perspectives adapted to the needs of the times and the concept of dual-creation in the teaching of specialized courses. To do a good job, it is necessary to make good use of the tools. Therefore, it is necessary to build a program for the construction of computer specialty teachers, explore ways and means to improve the theoretical and technical level of teachers, make full use of the excellent achievements in the field of AI, and encourage computer specialty teachers to participate in the training through a variety of ways. For teachers of computer professional experiment teaching, their teaching experience should not only be enriched but also have professional knowledge, and their skill level should be high. And in the computer professional experiment also guides students on how to use the relevant theoretical knowledge and lets the students understand the basis of gradual mastery, exercising students’ hands-on ability, knowledge, and skills. The teacher also has to have a certain sense of development to promote the development of students.
Facing the background of today’s artificial intelligence era, on the one hand, we need to improve the course evaluation mechanism to ensure the equality of education and teaching of computer professional courses. On the other hand, we need to adopt a combination of process and assessment of artificial intelligence teaching evaluation model. The process of teaching evaluation can be used with the classroom teaching evaluation, as well as a “fine course platform” on the data evaluation mode for the final assessment of the assessment of the students, the regular teaching evaluation, including attendance, online learning, the subject of the question and answer, the completion of the PTA on the computer homework, and so on. A good assessment mechanism, to a certain extent, in the teaching of computer specialties, constantly improves the status of teaching, and at the same time, the scientific assessment mechanism lets the students realize the importance of computers and causes them to pay attention to the computer laboratory teaching, the student’s enthusiasm for learning to stimulate, and promote their active participation in learning. Its teaching assessment mechanism in the implementation of the reform process, we must focus on students’ lifelong learning and the development of student personality, but also in line with the potential of students to refine so that students continue to develop, and then let the students learn to generate a certain momentum, and correct the correct mindset, to participate in computer laboratory teaching, to learn the relevant knowledge and skills.
On the one hand, according to the teaching characteristics of computer specialties and the characteristics of the teaching content, the construction of an open experimental teaching system, in order to realize the reform of the teaching mode to lay a good foundation, which should be clear about the training objectives and training direction of computer specialties, to understand the specific requirements of social positions on the talents, and to make full use of the cloud computing, artificial intelligence, etc., to meet the needs of different types of experiments, so as to effectively improve the quality of teaching. The experimental system mainly includes the following aspects: (1) cloud computing, mainly involving cloud storage, virtualization technology, cloud services, and other technologies. (2) Artificial Intelligence, which mainly involves Hadoops+HDFS technology, advanced analysis and model construction of artificial intelligence, data mining, and so on. On the other hand, relying on the cloud computing platform to build an information-based experimental teaching platform to enhance the openness of experimental teaching, such as the realization of the application of Hadoops+HDFS, which can be used to realize a number of functions on the teaching platform, such as teaching analysis, teaching management, and storage of teaching content, etc. Students can use their computers to log in to the virtual experimental teaching platform, which guarantees that students are able to intuitively understand the content of the experimental teaching and timely Communicate with teachers, analyze and discuss experimental resources and experimental results.
Computer-specialized experimental teaching under the background of artificial intelligence should focus on the stimulation of students’ subjective initiative, always adhere to the student-oriented concept in the teaching process, and give students the right to take the initiative, for example, let the students freely formulate the experimental subject, explore the content of their want to understand, and encourage students to express their views in the experiment as well as innovative practice, so as to make the whole process of experimental teaching more flexible and rich, so as to strengthen the cultivation of students’ ability. Cultivation of students’ ability.
In this paper, the modularized design of a university experimental teaching system based on an intelligent experimental platform is divided into two modules, namely, intelligent teaching software and intelligent assessment model, which are developed separately, and the structure of the experimental teaching system is shown in Figure 2. The intelligent teaching software is installed on the tablet computer of the intelligent experiment platform. Students log in to the software to carry out experimental course learning and experimental operations in electronic information and communication. The intelligent teaching software realizes the functions of intelligent face recognition, diversified contents of experimental courses, dynamic update of experimental courses, automatic acquisition of experimental data, automatic generation of experimental reports, and intelligent experimental assistant. After the experiment is completed, the software uploads the generated experiment report to the server. The intelligent evaluation model collects and processes the students’ experimental data in the experiment report and builds a BP neural network model to train and learn from these data. Through the continuous optimization of the intelligent assessment model, the function of predicting students’ experimental grades was realized.
The structure of the experimental teaching system
The characteristics of the university experimental teaching system based on an intelligent experiment platform are diversified experimental teaching, informatization of experimental process, and intelligent experimental evaluation.
Diversification of experimental teaching is reflected in the fact that the system includes a variety of teaching methods, including video teaching, Q&A teaching, and “experimental assistant”, in addition to traditional courseware teaching, to help students complete experiments in a multi-angle and omnidirectional way, and to improve students’ hands-on practical ability and independent problem-solving ability.
The informatization of the experimental process is reflected in the fact that from the beginning to the end of the experiment, all the experimental processes, including login, face recognition, completion of the experiment, and the generation of the final experimental report, are carried out on the intelligent teaching software of the intelligent experimental platform. On the one hand, it can realize the informatization management of students’ experimental reports, and on the other hand, it can collect other data in the process of students’ experiments as the reference of students’ performance evaluation while ensuring the authenticity of experimental data.
The intelligence of experimental assessment is reflected in the fact that this system utilizes the BP neural network of machine learning to construct an intelligent assessment model to assess students dynamically. Teachers can use the assessment results to intervene in advance to supervise students’ learning. Students can know their experimental course learning in advance by checking the assessment results, which improves students’ subjective initiative.
This study was conducted to apply an artificial intelligence-oriented teaching model in a computer science program at a university. Data such as computer literacy assessment scales of the experimental class (the teaching model of this paper) and the control class (the traditional method) were analyzed. There was no significant difference between the two classes before the experiment, and at the end of the experiment, the computer literacy assessment scale and computer knowledge test questions (post-test) were distributed again to the students in the two classes. Using the SPSS26 statistical tool, the post-test data of the computer literacy assessment scale of the two practical classes were tested for normality using the S-W test, and the P-value of the post-test data of the experimental class and the control class was greater than 0.05, so the original hypothesis that the post-test data of the two practical classes satisfy the normal distribution was not rejected.
In order to determine the differences in computer literacy between the control class and the experimental class after the experiment, the post-test data of computer literacy and the dimensions (computer awareness, computer knowledge, computer competence, computer affective and ethical) of the students in the two classes were subjected to an independent samples t-test, and the results of the data analysis are shown in Figure 3. In the dimension of computer awareness, the mean value of the experimental class is 26.834 and the control class is 20.842, and the difference between the two means. In the dimension of computer knowledge, the mean values of the experimental class and the control class are 33.358 and 21.936, respectively. In the dimension of computer competence, the mean values of the experimental class and the control class are 38.321 and 24.863, respectively. The difference between the two means is 13.458, which shows that the experiment has achieved some results in the cultivation of the computer competence dimension for the experimental class students. In the dimension of computer effectiveness and ethics, the mean value of the experimental class is 37.875, the mean value of the control class is 26.189, and the difference between the means of the two experimental classes is 11.686.
Data analysis results
The results of the independent samples t-test are shown in Table 1.
Test of independent sample t of computer literacy
| Ratio group (experimental class - comparison class) | Mean difference | Standard error difference | The difference is a 95% confidence interval | t | df | P | |
|---|---|---|---|---|---|---|---|
| Upper limit | Lower limit | ||||||
| YS1 | 3.018 | 0.375 | 2.320 | 3.84 | 8.314 | 89.143 | 0.001 |
| YS2 | 2.632 | 0.420 | 1.789 | 3.468 | 6.274 | 86.046 | 0.028 |
| YS3 | 0.415 | 0.440 | -0.487 | 1.296 | 0.721 | 87.576 | 0.074 |
| Computer consciousness | 6.075 | 0.740 | 4.689 | 7.563 | 8.326 | 89 | 0.001 |
| ZS1 | 2.926 | 0.436 | 2.032 | 3.812 | 6.742 | 89 | 0.007 |
| ZS2 | 2.936 | 0.422 | 2.163 | 3.816 | 7.184 | 85.346 | 0.001 |
| ZS3 | 3.185 | 0.403 | 2.496 | 3.972 | 7.836 | 89.942 | 0.001 |
| ZS4 | 2.463 | 0.395 | 1.695 | 3.225 | 6.285 | 78.826 | 0.063 |
| Computer knowledge | 11.532 | 1.236 | 9.022 | 13.922 | 9.342 | 70.725 | 0.002 |
| NL1 | 3.125 | 0.452 | 0.989 | 1.228 | 6.985 | 89 | 0.001 |
| NL2 | 1.584 | 0.487 | 0.442 | 0.552 | 7.362 | 82.912 | 0.063 |
| NL3 | 3.143 | 0.436 | 0.203 | 0.521 | 8.012 | 84.526 | 0.004 |
| NL4 | 3.594 | 0.502 | 0.134 | 0.143 | 6.284 | 85.923 | 0.002 |
| Computer capability | 13.436 | 1.704 | 10.063 | 16.795 | 8.023 | 82.232 | 0.002 |
| QY1 | 4.342 | 0.482 | 3.495 | 5.242 | 9.126 | 84.625 | 0.022 |
| QY2 | 1.643 | 0.642 | 1.309 | 2.936 | 2.463 | 89 | 0.017 |
| QY3 | 1.558 | 0.635 | 0.302 | 2.824 | 2.362 | 89 | 0.015 |
| QY4 | 4.236 | 0.473 | 3.311 | 5.147 | 11.326 | 89 | 0.001 |
| Computer Love and ethics | 11.725 | 1.633 | 9.432 | 14.925 | 7.842 | 86.843 | 0.001 |
| Computer literacy | 42.675 | 3.546 | 36.022 | 50.327 | 12.863 | 86.174 | 0.001 |
1) The p-value of the dimension of human-computer synergy awareness (YS3) is 0.074, which is greater than 0.05, indicating that there is no significant difference between the control class and the experimental class on the dimension of human-computer synergy awareness after the experiment. Whereas, the value consciousness (YS1) and exploration consciousness (YS2) are both less than 0.05, which rejects the original hypothesis and indicates that there is a significant difference between the two practicing classes on these two dimensions. The weaker performance of students in human-computer synergy awareness may be attributed to the fact that no attention was paid to the awareness of the relationship between artificial computers and society in the teaching and learning activities of the experimental class.
2) The p-values of basic knowledge of artificial computers (ZS1), application of artificial computer technology (ZS2), and conceptual principles of artificial computers (ZS3) are all less than 0.05, indicating that there is a significant difference between the students in the experimental class and the control class in these three dimensions compared to the control class, which suggests that the perceptual stage of artificial computers in the instructional model sets up the students’ personal experience in perceiving the relationship between the application of artificial computers and the autonomous Performing the activity of dissecting the principles of artificial computers helps students to have a deeper understanding of the conceptual principles of the subject of artificial computers, which is consistent with the results of Marseille’s study that found that students can better understand the concepts of artificial computers in the experiential and active inquiry sessions. The p-value of 0.063 for the dimension of the impact of artificial computers on humans (ZS4), on the other hand, indicates that the development of students in the experimental class on this dimension was not realized in the experiment. The possible reason for this is due to the fact that the specific knowledge of the relationship between artificial computers and human society was not carried out in the teaching process, thus preventing the students from correctly understanding and grasping the knowledge content of the specific impacts of artificial computers on various areas of society.
3) Further analyzing the significance, the p-value of Computational Thinking Ability (NL1), Programming Ability (NL3), and Artificial Computer Application Ability (NL4) is less than 0.05, which indicates that there is a significant difference between the students of the experimental class and the students of the control class in these three ability dimensions. The mean difference of the data competence (NL2) dimension is 1.584, and the p-value of 0.063 is greater than 0.05, which indicates that the experimental class had an improvement in data competence after the experiment compared with the control class, but the effect is not significant. The reason for this experimental result may be that, on the one hand, the students did not realize the importance and necessity of data for the field of artificial computing, and on the other hand, the students could not effectively obtain the required data or make decisions about the data. At the same time, students were not effectively guided to judge and use data in teaching activities. It is proved that the development of students’ data competence needs to be carried out in the task of acquiring and inputting data by students themselves, and therefore, the teaching activity was not emphasized in the course of this experiment, which resulted in the ineffective development of students’ data competence.
4) Further analysis of computer acceptance (QY1), computer fairness attitude (QY2), computer imagination (QY3), and computer ethics (QY4), the mean difference is 4.342, 1.643, 1.558, 4.236, respectively, in which the effect of the two dimensions of computer acceptance and computer ethics is more obvious. From the results of the independent samples t-test, the p-value of computer affective and ethical dimensions is less than 0.05, indicating that the difference between the experimental class and the control class in this dimension after the experiment is significant. This result indicates that having computer discursive activities in the teaching of computer courses can effectively cultivate students ’ attitudes of dialectical view of computers, have correct ethical concepts of computer application, and form good computer emotions.
In order to determine the effectiveness of the AI experimental teaching mode, a questionnaire based on the AI experimental teaching mode was prepared, which was designed in three dimensions: learning satisfaction, comprehensive ability enhancement, and collaborative communication. The comprehensive ability in the questionnaire includes practical ability, independent learning ability, problem-solving ability, teamwork ability, and knowledge acquisition ability. Knowledge acquisition, practical ability, and independent learning ability are in line with the teaching objectives of the validation-based experimental teaching mode, and problem-solving, practical ability, and teamwork are in line with the teaching objectives of the design-based experimental teaching mode. The object of the questionnaire survey is for the junior students of computer science majors in a university, and the implementation of the artificial intelligence-based experimental teaching mode is carried out in this class. 45 questionnaires were issued, and 45 questionnaires were statistically recovered, with a recovery rate of 100%.
The questionnaire has 12 questions divided into three dimensions, and the results obtained are now analyzed. First, comprehensive ability improvement dimension analysis: the results of the questionnaire are shown in Figure 4, 1~5: very satisfied, satisfied, average, dissatisfied, very dissatisfied. It can be seen through the statistical analysis of the students’ comprehensive ability improvement data. 83.7% of the students think that teaching in this teaching mode can improve their comprehensive ability. However, a small number of students disagree. The reason is that the students think that the teaching mode can not meet the needs of different students, which will affect their learning experience and the improvement of their comprehensive ability.
Questionnaire for teaching mode
Through the statistical analysis of the student’s satisfaction with this teaching experiment, it can be concluded that 86.6% of the students like the classroom method used in this teaching experiment have a strong interest in artificial intelligence experimental teaching, and think that the teaching mode is more interesting compared with the traditional teaching mode. Therefore, it can be concluded that students are very satisfied and interested in teaching based on the artificial intelligence experimental teaching mode adopted in this teaching experiment. However, there are a small number of students who are not satisfied with the teaching mode due to the fact that project-based teaching requires more time investment and students feel that the time is not enough to complete the task, and the sense of time urgency will lead to anxiety and reduce their satisfaction with the teaching mode.
86.4% of the students think that teaching in this teaching mode can cultivate their teamwork ability, and learning will be more active with cooperative group inquiry in artificial intelligence experiments. It can be seen that the teaching mode based on artificial intelligence experiments in colleges and universities can cultivate students’ teamwork ability, and by discussing and interacting with group members, they can jointly explore problems and share their views and experiences, thus promoting each other’s deepened understanding and improved learning effects. However, a small number of students are dissatisfied with the ability to improve collaborative communication because some students prefer independent learning, and project-based teaching usually emphasizes teamwork, which is less adaptable to those students who prefer to work independently or need a more personalized learning style.
The results of the survey are shown in Figure 5.82.2% of the students think that the experimental resources provided by the resource platform can meet the needs of learning, 17.8% think that it is average., 85.4% of the students think that the resource platform can make it more convenient to access the experimental resources, 11.2% think that it is average, and 3.1% don’t. 92.3% of the students think that the use of the resource platform can be a better choice to complete experimental tasks, 1.8% thought it was average, and 5.9% disagreed.87.1% of the students thought that using the resource platform to download was a better choice, 7.4% thought it was average and 5.5% disagreed.82.6% of the students thought that the resource platform provided sufficient experimental resources, 12% thought it was average and 5.4% disagreed.
Survey results
It can be seen that the experimental resources provided by this resource platform are helpful for students’ learning, and teachers can provide more abundant resources about AI experiments again. This resource platform can provide experimental resources for students more conveniently, and the experimental resources provided by this resource platform can help most of the students complete the experimental tasks, but some areas need to be improved. The experimental resources provided by this resource platform are helpful for students to carry out experiments on artificial intelligence, but there are also fewer resources, which can be increased to facilitate users to better find the required experimental resources.
This paper realizes the innovation of teaching mode through artificial intelligence technology, optimizing the curriculum system, practical links, faculty, and course intelligent evaluation mechanism. In the process of reforming the experimental teaching of computer majors, an intelligent teaching experimental system is constructed based on artificial intelligence technology, BP neural network, etc. After analyzing the teaching practice, it can be seen that there is a significant difference between the average scores of the experimental class and control class in the dimensions of computer awareness, computer knowledge, computer ability, computer sentiment and ethics, and computer literacy in which the average scores of computer ability and computer sentiment and ethics are improved significantly, and students’ overall academic level has been improved. There is a significant difference between the mean scores of the experimental class and the control class in the dimensions of computer knowledge, computer competence, computer emotions and ethics, and computer literacy, with computer competence and computer emotions and ethics being significantly improved and the overall academic level of the students being improved. Students’ satisfaction with the teaching mode and experimental resources exceeds 80%, and most of them believe that the teaching mode can cultivate their teamwork ability and that the experimental resources provided by the resource platform can meet their learning needs. Therefore, the AI-oriented teaching mode and experimental teaching in this paper can effectively improve students’ computer literacy level and academic performance.
This research was supported by the Collaborative detection and early warning platform for cyber espionage-type network attacks on military equipment (CXY2023013); Hebei Province’s first-class course “Principles of Computer Networks” (KC-2020-005-S).